a new converter topology for high-speed high-starting ...a new converter topology for high-speed...
TRANSCRIPT
A New Converter Topology for High-Speed
High-Starting-Torque Three-Phase Switched
Reluctance Motor Drive System
By
Ehab Elwakil
Submitted in partial fulfilment of the requirements for the degree of
Doctor of Philosophy
Department of Electronics and Computer Engineering
School of Engineering and Design
Brunel University
London, UK
January, 2009
ii
Acknowledgements
My full gratitude and thanks are to Allah the one who helped me and
gave me life, power and time to finish this work.
I am also grateful to my supervisor, Dr Mohamed K. Darwish for his
outstanding and continuous help, advice and support.
Also, I would like to pass my special gratitude to the soul of my father,
to my mother, my wife and kids for their patience and their emotional support
along the time of conducting this research.
All thanks are to the Egyptian government represented by the Egyptian
Cultural and Education Bureau for sponsoring this research in full, and to my
respectable teachers and colleagues in the department of power electronics
and energy conversion in the national Electronics Research Institute (ERI) in
Egypt for nominating me for this scholarship.
iii
Abstract
Switched reluctance motor (SRM) has become a competitive selection
for many applications of electric machine drive systems recently due to its
relative simple construction and its robustness. The advantages of those
motors are high reliability, easy maintenance and good performance. The
absence of permanent magnets and windings in rotor gives possibility to
achieve very high speeds (over 10000 rpm) and turned SRM into perfect
solution for operation in hard conditions like presence of vibrations or
impacts. Such simple mechanical structure greatly reduces its price. Due to
these features, SRM drives are used more and more into aerospace,
automotive and home applications. The major drawbacks of the SRM are the
complicated algorithm to control it due to the high degree of nonlinearity, also
the SRM has always to be electronically commutated and the need of a shaft
position sensor to detect the shaft position, the other limitations are strong
torque ripple and acoustic noise effects.
The aim of this research is to design a new drive circuit topology of the
SRM to meet the requirement of high-speed and/or high-starting torque
applications.
To accomplish this aim, the following objectives have been conducted:
1. A survey of the fundamental background of the SRM design,
applications, and theory of operation was established.
2. All converter topologies which have been used as an SRM drive
circuits in the literature were critically reviewed, compared and
classified according to their major field of application.
3. Upon completion of the literature review, decision was made to
select the asymmetric converter topology as it is widely used in
different applications and to modify the converter topology by
inserting the switched capacitance circuit which acts as a
variable capacitor and utilising this feature to profile the phase
current waveform.
iv
4. Two topologies of the switched capacitance circuit have been
analysed, mathematically modelled, and experimentally verified
to examine the behaviour of the switched capacitance circuit as
a variable capacitor.
5. A survey has been conducted on the software packages which
are commonly used to model and simulate the switched
reluctance machine and the Magnet software was selected for
its capability to model the motor with its real mechanical
dimensions and utilising the FEM analysis to evaluate the
magnetic characteristics of the motor besides the PSPICE-
complying models of the passive and power electronic
components of its circuit modeller.
6. A 3 phase 6/4 switched reluctance motor has been modelled
and the drive circuit was designed by modifying the conventional
asymmetric converter by inserting the switched capacitance
circuit within the converter to act as a variable capacitor
connected in series with the motor phase.
7. Simulation results were obtained and a comparison was held
between the conventional asymmetric converter results without
using the switched-capacitance circuit and the results obtained
after modifying the converter by inserting the switched-
capacitance circuit. The effect of the modification on different
parameters of the drive circuit was discussed and criticised.
The contribution to knowledge of this research in the field of SRM
drive circuits is represented in the following points:
• The Literature Review, which represents a
comprehensive critical review of the machine design and
its converters, which has been published as a journal
article in 2006, such a review is not established in the
literature up to the date of writing this thesis.
• introducing a new converter topology with a new
switching strategy utilising the switched capacitance
circuit to modify the conventional asymmetric converter
v
which is considered as a new topology of the converters
used to drive such a type of electric motors which has not
been covered in any of the literature published in this
topic.
• Developing an SRM model using Magnet software which
was not found in the literature.
• The contribution to solve the problem of torque ripples
and acoustic noise through phase current shaping without
varying the dwell angle.
• The contribution to present a new design which is perfect
for applications where high speed and/or high starting
torque is required.
The simulation results were very promising and it opens the
door for more research to be conducted to enhance the
performance of the suggested topology and to introduce the
concept of utilising the switched capacitance circuit to modify
other converter topologies of the SRM drive systems in the
future.
vi
Table of Contents
Acknowledgements ............................................................................................................. ii
Abstract .............................................................................................................................. iii
Table of Contents ................................................................................................................vi
List of Tables.......................................................................................................................ix
List of Figures ..................................................................................................................... x
List of Symbols .................................................................................................................. xiii
List of Abbreviations ...........................................................................................................xv
CHAPTER 1 ................................................................................................... 1
Introduction ......................................................................................................................... 1
1.1 Preface ......................................................................................................................... 1
1.2 SRM Applications .......................................................................................................... 3 1.2.1 Low Power Drives .................................................................................................. 3 1.2.2 Medium Power Drives ............................................................................................ 3 1.2.3 High Power Drives .................................................................................................. 4 1.2.4 High Speed Drives ................................................................................................. 4 1.2.5 Emerging Applications ............................................................................................ 4 1.2.6 High volume Applications ....................................................................................... 5 1.2.7 Underwater Applications ......................................................................................... 5 1.2.8 Linear Drive Applications ........................................................................................ 5
1.3 Aim of this Research ..................................................................................................... 6
CHAPTER 2 ................................................................................................... 8
Literature Review ................................................................................................................ 8
2.1 Introduction ................................................................................................................... 8
2.2 Configurations of Switched Reluctance Motor ................................................................ 9
2.3 Converter Topologies for Switched Reluctance Motor Drives ....................................... 12
CHAPTER 3 ................................................................................................. 25
Fundamentals and Theory of Operation of SRM ................................................................ 25
3.1 Motor Construction ...................................................................................................... 25
3.2 Theory of Operation..................................................................................................... 26
3.3 Torque Production ....................................................................................................... 27
vii
3.4 Equivalent Circuit of a SRM ......................................................................................... 30
3.5 Static Characteristics of the SRM ................................................................................ 34 3.5.1 Flux Linkage Current Curves ................................................................................ 34
3.5.1.1 Aligned rotor position ..................................................................................... 34 3.5.1.2 Unaligned rotor position ................................................................................. 35 3.5.1.3 Flux linkage current curves at intermediate rotor positions ............................. 36
3.6 Torque-Speed Characteristics ..................................................................................... 39
3.7 Control Strategies of Switched Reluctance Motor Drives .............................................. 40 3.7.1 Speed Control ...................................................................................................... 41 3.7.2 Current Control ..................................................................................................... 41 3.7.3 Torque Control ..................................................................................................... 42
3.8 Control Principle of SRM ............................................................................................. 43
3.9 Sensor-less Operation of SRM Drives.......................................................................... 46 3.9.1 Incremental Inductance Measurement [1] ............................................................. 47 3.9.2 Observer based rotor position estimation .............................................................. 47 3.9.3 Inductance Sensing .............................................................................................. 47
3.9.3.1 Phase Pulsing ............................................................................................... 48 3.9.3.2 Frequency Modulation ................................................................................... 48 3.9.3.3 Phase Modulation.......................................................................................... 48 3.9.3.4 Amplitude Modulation .................................................................................... 48 3.9.3.5 Self-Voltage Technique ................................................................................. 49 3.9.3.6 Artificial-Intelligence-based inductance estimation [58]................................... 49
CHAPTER 4 ................................................................................................. 59
Switched Capacitance circuit applied to an SRM Driver topology ....................................... 59
4.1 Proposed technique to profile the phase current .......................................................... 59
4.2 Switched capacitance circuit ........................................................................................ 59 4.2.1 Single Capacitance, Double Switch (SCDS) Circuit............................................... 60
4.2.1.1 The Switching Function ................................................................................. 62 4.2.2 Double Capacitance, Double Switch (DCDS) Circuit ............................................. 66
4.3 Implementation of the Switched-capacitance Circuit .................................................... 71
4.4 Other topologies of switched capacitance circuit .......................................................... 75
4.4.1 Single-capacitance triple-switch circuit ...................................................................... 76
4.4.2 Single-capacitor 5-switches circuit ............................................................................ 77
4.4.3 Hybrid switched capacitance systems ....................................................................... 77
4.5 Applying the switching capacitance circuit to SRM drive topology ................................ 79
CHAPTER 5 ................................................................................................. 80
Modelling and Simulation of the Proposed Drive Circuit ..................................................... 80
5.1 Introduction ................................................................................................................. 80
5.2 Magnet Software ......................................................................................................... 81
viii
5.3 Parameters of the Simulated Machine ......................................................................... 83
5.4 Drive Circuit................................................................................................................. 84 5.4.1 Asymmetric Converter Drive ................................................................................. 85
5.4.1.1 Simulation results for the Asymmetric converter drive .................................... 86 5.4.2 Single Capacitance Double Switch (SCDS) drive Circuit ....................................... 89
5.4.2.1 Switching Strategy......................................................................................... 89 5.4.2.2 Calculation of the effective capacitance, Ceff .................................................. 90 5.4.2.3 Calculation of Duty Cycle, D .......................................................................... 90 5.4.2.4 Determining the value of the fixed capacitor, C .............................................. 91 5.4.2.5 Simulation results for SCDS drive circuit ........................................................ 91
(a) Effect of SCC on the phase current profile ....................................................... 93 (b) Effect of SCC on the developed torque ............................................................ 95 (c) Effect of SCC on the voltage stresses .............................................................. 96 (d) Effect of SCC on the current harmonics ........................................................... 97 (e) Effect of the SCC on the overall efficiency ....................................................... 99 (f) Effect of the SCC on the switching losses ....................................................... 101
5.4.2.6 Effect of the Switching Frequency................................................................ 102 5.4.3 Two Phase Operation ......................................................................................... 111 5.4.4 Double-Capacitance Double-Switch Circuit ......................................................... 120
5.4.4.1 Determining the values of C1 and C2 ............................................................ 120 5.4.4.2 Calculation of the duty cycle ........................................................................ 121 5.4.4.3 Simulation results of the DCDS drive circuit ................................................. 122
CHAPTER 6 ............................................................................................... 124
Conclusions and Future Work ......................................................................................... 124
6.1 Conclusions............................................................................................................... 124
6.2 Future work ............................................................................................................... 128
References ..................................................................................................................... 130
Appendix ......................................................................................................................... 138
ix
List of Tables
Table 2-1 Comparison of Common SRM Drive Circuits .................................. 20 Table 3-1 Comparison of design parameters and objective functions. .......... 57
Table 5-1The machine parameters. .................................................................. 83
x
List of Figures
Figure 2.1 Basic Configurations of SRM. .......................................................... 10
Figure 2.2 Different implementations of Stepped- gap Switched Reluctance Motors. ............................................................................................. 11
Figure 2.3 Two-phase homopolar SRM. ........................................................... 11
Figure 2.4 Structure of c-core stator SRM. ....................................................... 12
Figure 2.5 Notched rotor teeth. .......................................................................... 12
Figure 2.6 Asymmetric Bridge Converter.......................................................... 14 Figure 2.7 R-Dump Converter............................................................................ 14
Figure 2.8 Bifilar Type Converter. ...................................................................... 15
Figure 2.9 Split DC Supply Converter. .............................................................. 15
Figure 2.10 C–Dump Converter......................................................................... 16
Figure 2.11 Variable DC Link Converter. .......................................................... 17
Figure 2.12 Resonant Converter. ...................................................................... 18
Figure 2.13 Two Stage Power Converter. ........................................................ 19 Figure 3.1 Basic Structure of SRM. ................................................................... 25
Figure 3.2 Single Element of SRM. ................................................................... 26
Figure 3.3 Ideal profile of phase inductance. ................................................... 27
Figure 3.4 Magnetization Curve......................................................................... 31 Figure 3.5 Single phase equivalent circuit of a SRM. ...................................... 34
Figure 3.6 Measured flux linkage-current curves at aligned and unaligned positions........................................................................................... 36
Figure 3.7 Flux linkage-current curves of 6/4 SRM. ........................................ 38
Figure 3.8 The three dimension of flux linkage, current, and position angle curves. ............................................................................................. 38
Figure 3.9 SRM Torque-Speed Characteristics. .............................................. 39
Figure 3.10 Speed Control Closed-loop Block Diagram. ................................ 41
Figure 3.11 Switching strategy for motor operation. ........................................ 43
Figure 3.12 Typical current waveforms for various advance and dwell angles. .......................................................................................................... 45
Figure 3.13 Input and output membership functions for fuzzy inductance estimation. ....................................................................................... 50
Figure 3.14 Phase inductance, (a) Experimentally obtained. (b) fuzzy-logic-based. .............................................................................................. 51
Figure 3.15 ANN-based position estimator....................................................... 52
Figure 3.16 ANFIS structure for 2 inputs and 1 output. ................................... 53
Figure 3.17 Structure of the ANFIS. .................................................................. 55
Figure 3.18 ANFIS models ................................................................................. 55
Figure 3.19 ANN-based force estimator. .......................................................... 56
Figure 3.20 Optimal design of SRM by GFA. ................................................... 57
Figure 4.1 Single Capacitance Double Switch Circuit. .................................... 60
Figure 4.2 Switched capacitance circuit with a current limiter. ....................... 61
Figure 4.3 Switching Function............................................................................ 62
Figure 4.4 Relation between D and Ceff for SCDS circuit. ............................... 66
Figure 4.5 Double Capacitance Double Switch Circuit.................................... 67 Figure 4.6 DCDS topology with a current limiter. ............................................. 67
xi
Figure 4.7 Relation between D and Ceff for DCDS circuit as referred to branch C1. ........................................................................................ 69
Figure 4.8 Relation between D and Ceff for DCDS circuit as referred to branch C2. ........................................................................................ 71
Figure 4.9 Schematic diagram of the implemented switched capacitance circuit. ............................................................................................... 72
Figure 4.10 Experimental setup of the SCC. .................................................... 72
Figure 4.11 (a) Drive circuit of the MOSFET switches. (b) Microprocessor board. ............................................................................................... 73
Figure 4.12 Experimental plot of D against Ceff for SCDS circuit. .................. 74
Figure 4.13 Experimental plot of D against Ceff for DCDS circuit. .................. 74
Figure 4.14 Single-capacitor triple-switches. .................................................... 76
Figure 4.15 Single-capacitor, triple-switches circuit with a coil. ..................... 76 Figure 4.16 Single-capacitor 5-switches circuit. ............................................... 77
Figure 4.17 Single-capacitor triple-switches system with SCC. ..................... 78
Figure 4.18 Single-capacitor 5-switches system with SCC. ........................... 78
Figure 4.19 Single-capacitor 6-switches system with SCC. ........................... 79 Figure 5.1 Geometric Modeller of Magnet software. ....................................... 81
Figure 5.2 Circuit modeller of Magnet software. .............................................. 82
Figure 5.3 Finite element mesh of the solved model. ...................................... 83 Figure 5.4 Proposed drive circuit. ...................................................................... 84
Figure 5.5 Conventional Asymmetric Converter. ............................................. 86
Figure 5.6 Phase current profile for single phase operation mode. ............... 87
Figure 5.7 Voltage across the motor phase. .................................................... 88 Figure 5.8 Torque pulse developed by a single phase.................................... 88
Figure 5.9 Drive circuit utilising single-capacitance double-switch circuit. .... 89
Figure 5.10 Phase current profile for SCDS circuit, f=1kHz............................ 92
Figure 5.11 Voltage across the motor phase for SCDS circuit, f=1kHz. ........ 92
Figure 5.12 Torque pulse developed by a single phase for SCDS circuit, f=1kHz. ............................................................................................. 93
Figure 5.13 Current profiles for the asymmetric converter with and without SCC. ................................................................................................. 94
Figure 5.14 Developed torque for the asymmetric converter with and without SCC. ................................................................................................. 95
Figure 5.15 Voltage stress over the machine phase. ...................................... 96 Figure 5.16 Voltage stress across the upper switch, S4. ................................. 97
Figure 5.17 Respective Harmonic Amplitudes. ................................................ 98
Figure 5.18 Total Harmonic Distortion (THD). .................................................. 98
Figure 5.19 Variation of THD with switching frequency. ................................. 99 Figure 5.20 Ohmic Losses per coil. ................................................................. 100
Figure 5.21 Instantaneous Magnetic Energy.................................................. 101
Figure 5.22 Voltage stresses over S1 and S2 at 1kHz. .................................. 102
Figure 5.23 Current profiles for 300 Hz and 500 Hz. ..................................... 103
Figure 5.24 Flux linkage at 1.5 kHz. ................................................................ 104
Figure 5.25 Current profiles for 1kHz and 1.5 kHz. ....................................... 105
Figure 5.26 Current profiles for 2 kHz, 2.5 kHz and 3 kHz. .......................... 105 Figure 5.27 Current profiles for 4 kHz and 5 kHz........................................... 106
Figure 5.28 Current profiles for 5.5 kHz and 6 kHz. ...................................... 106
Figure 5.29 Current profiles for 6.5 kHz and 7 kHz. ...................................... 107
Figure 5.30 Current profiles for 8.5 kHz and 10 kHz. .................................... 108
xii
Figure 5.31 Current profiles for 20 kHz and 30 kHz. ..................................... 108
Figure 5.32 Current profiles for 40 kHz and 50 kHz. ..................................... 109 Figure 5.33 Current profiles for 60 kHz and 70 kHz. ..................................... 109
Figure 5.34 Current profiles for 80 kHz and 100 kHz. ................................... 110
Figure 5.35 Switching losses in terms of the switching frequency. .............. 111
Figure 5.36 Phase Currents in case of overlap, without SCC. ..................... 113
Figure 5.37 Developed torque in case of overlap, without SCC................... 113
Figure 5.38 Phase Currents in case of overlap, using SCC. ........................ 114
Figure 5.39 Developed torque in case of overlap, using SCC...................... 114 Figure 5.40 Torque ripples in asymmetric converter. .................................... 115
Figure 5.41 Developed torque at f=30 kHz. .................................................... 116
Figure 5.42 Developed torque at f=40 kHz. .................................................... 116
Figure 5.43 Developed torque at f=70 kHz. .................................................... 117 Figure 5.44 Developed torque at f=80 kHz. .................................................... 117
Figure 5.45 Developed torque at f=100 kHz................................................... 118
Figure 5.46 Flux Linkage versus Stator Current (Simulated). ...................... 119
Figure 5.47 maximum increase in output energy at 20 kHz. ........................ 119
Figure 5.48 Double-Capacitance Double-Switch Circuit. .............................. 120
Figure 5.49 DCDS circuit for f=1 kHz. ............................................................. 122
Figure 5.50 DCDS circuit for f=10 kHz. ........................................................... 123
xiii
List of Symbols
θ Rotor angle (position)
φ Flux linkage
ω Angular speed
ϕ Phase angle
ψ Phase shift
γ C2/C1
τ Switching interval period
Λ Tooth pitch
δ Air gap length
θa Advance angle
θal Aligned position
θco Commutation angle
θd Dwell angle
∆i Hysteresis current window
θomax Maximum overlap angle
βr Rotor pole arc
θr Rotor position
θrn Normalised rotor position
βs Stator pole arc
ωs Switching angular frequency
φs Stator flux linkage
∆T Time period
θu Unaligned position
Ao, An, Bn Fourier coefficients C Capacitance Ceff Effective capacitance D Duty cycle Dr Rotor external diameter Dri Rotor internal diameter Ds Stator external diameter Dsi Stator internal diameter Dw Winding-wire diameter f frequency F Magneto-motive force f(t) Switching function f’(t) Complementary switching function fs Switching frequency g Air gap width hr Rotor tooth-height hs Stator tooth-height i Instantaneous phase current
I^ Peak value of current
i* Current command signal iac Phase ac current
xiv
Im Maximum current Ip Peak flat tip of current signal Iph RMS phase current is Stator current Kb EMF constant L Inductance La Aligned inductance Las Aligned saturation inductance Lcs Stator-core axial length Lph Phase inductance Lu Unaligned inductance Nph Number of turns per stator-phase Pr Number of rotor poles q Number of phases R Resistance Rm Motor phase resistance S Switch T Torque T Period toff OFF time ton ON time tr Rotor tooth width ts Stator tooth width us Stator voltage v Instantaneous voltage V* Converter command signal vac Phase ac voltage vc Capacitor voltage Vm Maximum voltage Vph RMS phase voltage vs Supply voltage
v xx
^
Peak value of voltage
Wc Magnetic co-energy Wf Magnetic field energy Wm Mechanical energy Xc Capacitive reactance XL Inductive reactance Xph Phase reactance yr Rotor internal radius ys Stator steel thickness Zph Phase impedance
xv
List of Abbreviations
AI Artificial intelligence ANFIS Adaptive Neuro-fuzzy inference system ANN Artificial neural network DCDS Double-capacitance, double-switch EMF Electromotive force EMI Electromagnetic interference FEM Finite element mesh GFA Genetic fuzzy algorithm MF Membership function MMF Magnetomotive force PID Proportional+ integral+ derivative PWM Pulse width modulation SCC Switched capacitance circuit SCDS Single-capacitance, double-switch SRM Switched reluctance motor TR Torque ripple ZCS Zero current switching ZVS Zero voltage switching PIC Programmable interface controller
in some literatures referred to as programmable intelligent computer
DSP Digital signal processor
1
Chapter 1
Introduction
1.1 Preface
The switched reluctance motor (SRM) represents one of the earliest
electric machines which was introduces two centuries back in the history. It
was not widely spread in industrial applications such as the induction and DC
motors due to the fact that at the time when this machine was invented, there
was no simultaneous progress in the field of power electronics and
semiconductor switches which is necessary to drive this kind of electrical
machines properly. The problems associated with the induction and DC
machines together with the revolution of power electronics and
semiconductors late in the sixties of the last century (1969), led to the re-
invention of this motor and redirected the researchers to pay attention to its
attractive features and advantages which help with overcoming a lot of
problems associated with other kinds of electrical machines such as; brushes
and commutators in DC machines, and slip rings in wound rotor induction
machines, besides the speed limitation in both kinds. The simple design and
robustness of the switched reluctance machine made it an attractive
alternative for these kinds of electrical machines for many applications
recently, specially that most of its disadvantages which are mentioned in the
following chapter could be eliminated or minimized using the high speed and
high power semiconductor switches such as the power Thyristors, power
GTO, power transistors, power IGBT, power MOSFET. The availability and
the inexpensive cost of these power switches nowadays besides the
presence of microcontroller, microprocessors, PIC and DSP chips makes it a
strong opponent to other types of electrical machines.
In industry, there is a very wide variety of designs of the switched
reluctance machines which are used as motors or generators, these designs
vary in number of phases, number of poles for both stator and rotor, number
2
of teeth per pole, the shape of poles and whether a permanent magnet is
included or not.
These previous options together with the converter topology used to
drive the machine lead to an enormous number of designs and types of
switched reluctance machine systems, which means both the switched
reluctance machine with its drive circuit to suit different applications with
different requirements. It is to be noted that it is well known to those who are
interested in this kind of electrical machines that the drive circuit and the
machine is one integrated system, one part of such a system can’t be
separately designed without considering the other part.
A Switched Reluctance (SR) machine is a rotating electric machine
where both stator and rotor have salient poles. The stator winding is
comprised of a set of coils, each of which is wound on one pole. SR motors
differ in the number of phases wound on the stator. Each of them has a
certain number of suitable combinations of stator and rotor poles.
When operated as a motor, the machine is excited by a sequence of
current pulses applied at each phase. The individual phases are
consequently excited, forcing the motor to rotate. The current pulses need to
be applied to the respective phase at the exact rotor position relative to the
excited phase.
The inductance profile of SR motors is triangular shaped, with maximum
inductance when it is in an aligned position and minimum inductance when
unaligned. When the voltage is applied to the stator phase, the motor creates
torque in the direction of increasing inductance. When the phase is
energized in its minimum inductance position, the rotor moves to the
forthcoming position of maximal inductance. The profile of the phase current
together with the magnetization characteristics define the generated torque
and thus the speed of the motor.
There are several advantages of the switched reluctance machine that
gives it preference over other types of electrical motors in many applications,
these advantages are: [1-5].
1. Simple design and robust structure.
2. Unwind rotor.
3
3. Lower cost.[6]
4. High starting torque without the problem of inrush currents
compared with induction motor.
5. Suitable for high speed applications.
6. High reliability due to the electric and magnetic independency
of the machine phases.
7. Suitable for high temperature applications compared to other
machines of similar ratings.
8. Motor torque is independent of the phase current polarity.
9. Four quadrant operations.
10. A wide constant torque or power region in the torque speed
characteristics.
11. High efficiency throughout every part of torque speed range.
1.2 SRM Applications
Many of the applications are categorized as low power, medium power,
high power, and high speed drives for rotary motor drive. Some emerging
applications and underwater applications are shown as well as few
applications for linear switched reluctance motors. [1]
1.2.1 Low Power Drives
In this category, drives less than 3 hp are considered, this has been a
successful range for many applications of switched reluctance motors such
as, plotter drive, air handler motor drive, hand fork lift/pallet truck motor drive,
door actuator system, washers, and dryers.
1.2.2 Medium Power Drives
This segment is generally in the range of less than 300 kW for general
purpose, variable speed applications. New applications do not come in large
numbers and don’t create huge markets for the switched reluctance machine
drives in the medium power range. However, SRM drive systems have been
4
developed within the medium power range for general purpose industrial
applications (up to 140 hp), train air conditioning drives (40 kw), and mining
drive (300 kw) applications.
1.2.3 High Power Drives
There have been some efforts to develop SRM drives up to 1000 hp for
fan and pump applications. At this high power level, the SRM converter is
very competitive due to the fact that at high power levels, the inverter
switches do not come with anti-parallel diodes as a single package which
leads to a significant reduction in the converter cost added to the cheaper
cost of the SRM compared with other AC or DC drives.
1.2.4 High Speed Drives
The SRM is a natural choice for high speed applications due to its
robust rotor construction and its high power density. The performance
sensitivity to high temperatures in the SRM is minimal compared with the
permanent magnet synchronous and brushless dc motor and induction motor
drive systems. Some existing products are available in the market for high
speed applications such as screw rotary compressor drive at 4500 rpm,
centrifuge for medical applications with high speed at 30000 rpm, aircraft jet
engine starter motor generator at 27000 rpm, at the same time, the SRM can
be used in the regeneration mode to supply electric power to the aircraft, the
regeneration speed range is between 27000 and 50000 rpm. Also, fuel lube
pump motor drive for a gas turbine engine for is another aerospace
application of the SRM.
1.2.5 Emerging Applications
Variable speed applications can be found in hand power tools, fans,
pumps, drives for freezers and refrigerators, automotive applications such as
antilock brake drives, transportation and machine tool spindles. Also,
underwater applications for marine propulsion system for surface,
submersible, and flexible steering propulsion units. Finally, linear drive
5
applications such as direct drive based elevator door openers,
semiconductor fabrication plants for material handling, low speed direct
propulsion drive transit systems, automatic direct drive base door openers in
office and commercial environments, and slide tables.
1.2.6 High volume Applications
Due to its simplicity of construction and resulting economy in its
manufacture, SRMs are expected to gravitate toward high volume but
inexpensive variable speed applications, such as hand power tools, fans,
pumps, drives for freezers and refrigerators, automotive applications such as
antilock brake drives and ripple torque minimized electronically controlled
steering system and transportation and machine tool spindles where torque
ripple is not a serious concern.
1.2.7 Underwater Applications
This drive system is also under consideration for underwater propulsion
applications. The main features sought after for marine propulsion systems
for surface, submersible, and flexible-steering propulsion units are the
enhanced control, greater efficiency, reduced acoustic noise, high reliability,
and equipment layout flexibility. One such application involves a 7.5 kw drive
system with a partial stator and a rotor completely in water. The other
application is an underwater marine drive unit for propulsion of remotely
operated vehicles.
1.2.8 Linear Drive Applications
The linear switched reluctance motor is emerging for select
applications. The trend is to use direct drive systems whereby the rotary to
linear mechanical converters such as the lead screw, gears, etc., can be
dispensed with both to reduce the cost of the system and to make the
system compact and highly reliable. The linear switched reluctance motors
have been in long use in XY table drive systems. Also it is found in direct
drive based elevator door openers, semiconductor fabrication plants for
6
material handling, low speed direct propulsion drive transit systems,
automatic direct drive based door openers in office and commercial
environments, machine tools, slide tables, and in futuristic levitation systems
for urban transit systems.
1.3 Aim of this Research
The aim of this research is to design a new topology of SRM drive
circuit which is capable of shaping the phase current waveform in order to
minimize the rise time and the fall time which in turn leads to two significant
results, first, to achieve a more flat current pulse within the same pulse width
(the same dwell angle), hence increasing significantly the net developed
torque as the latter is proportional to the square of the phase current,
secondly, is to meet the condition of better performance at higher speeds.
The new topology is a modification of the conventional asymmetric
converter topology, Figure 2.6. The modification is achieved by inserting a
switched capacitance circuit in series with the motor phase as shown in
Figure 5.4.
The switched capacitance circuit in this case acts as a variable
capacitor which may be varied by changing the duty cycle of the two
semiconductor switches so that the value of the equivalent capacitance
tracks the variation of the phase inductance profile of the machine causing a
significant change in both the amplitude and the shape of the current
waveform.
The novelty of introducing the switched capacitance circuit to the field of
SRM drives is a main feature of this research. The main applications of the
switched capacitance circuits in electrical power field are the power active
filters and reactive power generation, while no research has been conducted
to introduce it to the field of SRM drive circuits.
7
The thesis consists of six chapters, broken down as follows:
Chapter 1 is an introduction of the aim and the break down of the
thesis.
Chapter 2 is a critical review of the SRM designs and the different
topologies of the SRM drive circuits with a critical comparison showing the
advantages and the disadvantages of each topology together with the
suitable applications for it. Also, a brief description of the different control
schemes is presented.
Chapter 3 presents the fundamentals and theory of operation of the
switched reluctance motor showing the mathematical equations that govern
the electrical, the magnetic and the mechanical characteristics of the
machine.
Chapter 4 shows the mathematical and analytical derivation of the
characteristics of two different topologies of switched capacitance circuits to
be considered in this research which are the single capacitance double
switch, and the double capacitance double switch circuits, deducing the
relation between the effective capacitance and the duty cycle of the
semiconductor switches.
Chapter 5 presents the modeling of a 3-ph, 6/4 switched reluctance
machine using the Magnet software and simulation results are presented and
discussed in details for each of the above two cases.
And finally the conclusions and future work are presented in chapter 6.
8
Chapter 2
Literature Review
2.1 Introduction
Switched reluctance motors (SRM) and their drives for industrial
applications are considered in many references to be of recent origin.
Krishnan, referred the first proposal of a variable reluctance motor for
variable speed application to 1969, and the originality of the motor itself to
1842 [1].
Some other authors dated back the first acknowledged application of
Switched Reluctance Motor to the late 1830s [7]. Since that date, the
Switched Reluctance Motor was proposed but with major drawbacks, which
severely limited the applications of this motor in industry, these major
drawbacks are:
- Necessity of electronic commutation [8].
- Complicated analysis due to non-linear characteristics.
- Difficulty of control.
- Need to shaft position sensor.
- Torque ripples.
- Acoustic noise.
However, if considering the fact that the Switched Reluctance Motor
together with its drive and control circuits should be designed as one single
block, it can be appreciated why the SRM was not applicable since the early
date on which it was invented; that is because of the unavailability of fast,
inexpensive, high-power switching devices and control techniques at that
time till the late 1960s, at which, most of the SRM drawbacks could be
overcome [7]. Many researchers paid attention to the analysis and modelling
of the SRM since that date, specially with the great development of software
9
packages, which helped with accurate modelling of the machine parameters
such as PSPICE and MATLAB—[9-30].
This fact is more appreciated when one knows that unlike DC or induction
motors, the Switched Reluctance Motor can’t run directly neither from a DC
nor an AC supply, instead a power converter must supply unipolar current
pulses, timed accurately according to the inductance profile to produce
positive torque in the motoring operation mode [31].
It is of great importance to overview the various designs of Switched
Reluctance Motors as both the motor design and its drive circuit are severely
inseparable.
2.2 Configurations of Switched Reluctance Motor
The Switched Reluctance Motor is a doubly salient, single excited
machine, which has a number of critical parameters should be considered in
its design process; these parameters are : the dimensions of stator and rotor
laminations, winding details, number of poles, pole arcs, and number of
phases[1]. These critical parameters determine the electrical, magnetic,
mechanical, and thermal capabilities of the Switched Reluctance Motor.
The variety of combinations of number of phases with stator and rotor
number and shapes of poles led to a wide range of possible designs of the
SRM. The Rotary Switched Reluctance Motors, which is the scope of this
research, could primarily be divided into Radial field and Axial field
machines, Figure 2.1.
Within the radial field rotary machines, there are two possible designs, if
the stator and the rotor of the machine are symmetrical about their centre
lines and equally spaced around their periphery, this may be referred to as a
regular structured machine as this term was first adopted by T. Miller [31].
10
Figure 2.1 Basic Configurations of SRM.
In the early 1990s, an alternative arrangement, called the dual rotor
pitch machine was introduced in the design of a five-phase 10/8 SRM but
with its rotor having two separate pole pitches, the authors introduced the
position theory as an extension to the basic modelling theory to encompass
the boarder range of SRMs [31]. The position theory added a new
categorization concept to SRMs on basis of the relative position of rotor
poles with respect to the nearest stator poles, according to which, machines
could be categorized to one position, two positions, or four positions machine
[31].
Researches related to the machine design have been carried out to
overcome the major drawbacks of SRM, specially those related to acoustic
noise and torque ripples. One other problem with the SRM arises only in
single and two phase machines, which is the existence of dead zones which
may lead to the failure of the motor to self start at certain positions of the
rotor.
The hybrid SRM was proposed to overcome this problem in the late
1980s, this type of motor utilizes a PM inserted either in the rotor or in the
stator iron in order to guarantee the possibility of providing starting torque at
all rotor positions. The idea of hybrid or PM SRM was applied later to multi
phase (3 and above) machines to enhance the torque/volume ratio and to
boost the efficiency of the machine [32, 33]. Other solutions were proposed
SRM
Rotary Linear
Radial field Axial field Longitu-dinal flux
Transv-ersal flux
Basic Structure
Short flux path
Single stack
Multi-stack
11
to solve the self starting problem such as, the stepped-gap motor, Figure 2.2
which was proposed in 1986 [34], and the homopolar SRM, which was
applied to a two phase SRM design as shown in Figure 2.3 [35].
Figure 2.2 Different implementations of Stepped- gap Switched Reluctance Motors.
The axial field SRM in which the flux path is parallel to the shaft was
proposed in the early 1970s to meet the applications in which the total length
may be constrained, such as ceiling fans. The axial field SRMs have the
advantage of lower core losses than the radial field motors, however, they
have higher mutual inductance and higher magnetic pull on the rotor [1].
Figure 2.3 Two-phase homopolar SRM.
There are various configurations of SRM designs which were proposed
to improve the overall performance of the machine, e.g. a c-core stator was
proposed in [36] as shown in Figure 2.4 to increase torque capability and
efficiency and also to provide a higher flexibility in winding design, while a
new pole tip shape was suggested in [37], and a new notched teeth rotor
shape was suggested in [38] to reduce the torque ripple rate as shown in
Figure 2.5.
12
Figure 2.4 Structure of c-core stator SRM.
In [4], a novel SRM with wound-cores which is made of grain-oriented
silicon steel was proposed to reduce the weight and volume of the SRM,
hence increasing the power/weight and power/volume capability.
Figure 2.5 Notched rotor teeth.
2.3 Converter Topologies for Switched Reluctance Motor Drives
It is essential before considering the different topologies of Switched
Reluctance Motor drive circuits to introduce the following features of the
motor which give it an attractive advance over other types of AC machines;
and affects the design of the drive circuit, these features are:
1. The torque in a SRM is independent of the polarity of the phase
excitation current, which means only one switch per phase winding is
required to energize or de-energize this phase.
13
2. There is always one switch connected in series with a phase winding,
which makes the SRM self protected against shoot-through fault.
3. The phases of the SRM are electrically and magnetically
independent, which allows separate control of current and torque for
each phase, and also allows possible operation of the machine in
case of one phase winding failure, although with reduced output
power.
4. The mutual coupling between phases in a SRM can be neglected,
this also gives an advance for independent phase current and torque
control.
5. The phase inductance of the SRM varies with the rotor position in
either a linear or a nonlinear profile.
The previous facts about the SRM explain the expected function of the
drive circuit for this type of AC machines, which could be concluded in three
tasks; first, is to divert current into the phase only in the positive gradient
period of its inductance profile. Second, is to shape the energizing current of
each phase, including its amount and its rise and fall times, of course less
rise and fall times are desired to maximize the torque productivity during
motoring operation[17, 39][8]. Third task is to provide a path for the stored
magnetic energy in the phase winding during the commutation period,
otherwise it will result in excessive voltage stress across the phase winding,
hence across the semiconductor switching element leading to its failure, this
energy could be freewheeled, or returned to the DC source either by
electronic or electromagnetic means.
There are several topologies suggested to achieve the above function
of the drive circuit. These topologies are well classified in [1] based on
the number of switches used to energize and commutate each phase.
The most common configuration of SRM drive circuit is the asymmetric
type with freewheeling and regeneration capability, Figure 2.6. The
performance of this converter depends on the switching strategy, which
varies between simultaneous switching of T1 and T2, or switching one switch
ON and OFF while the other is ON all time. Selecting the proper switching
strategy, dwell angle and control technique (usually hysteresis current
control) will define the efficiency and application of this converter.
14
The asymmetric converter is the most common for voltage source
operation for low power levels [40].
Figure 2.6 Asymmetric Bridge Converter.
Figure 2.7 R-Dump Converter.
15
Figure 2.8 Bifilar Type Converter.
The R-dump converter, Figure 2.7, is one of the configurations which
has one switch and one diode per phase. The value of R determines the
power dissipation and the switch voltage. A compromise of R value should
be done to achieve both reasonable stress on the switch (increases with
higher R), and appropriate fall time of the current (increases with lower R).
Analysis and detailed design procedure of this converter is explained in [41].
The Bifilar type converter, Figure 2.8, also has one switch and one
diode per phase but with the capability of regenerating the stored magnetic
energy to the supply by the bifilar winding. This converter is also utilizing one
switch per phase. The bifilar converter has the drawbacks of reduced power
density, and the high stress on the switching elements due to the bifilar
windings inductance.
Figure 2.9 Split DC Supply Converter.
16
Figure 2.9 shows the split DC supply converter which allows
freewheeling and regeneration. Proper design of the values of C1 and C2
should be done carefully to achieve charge balancing across the DC link.
However, this converter utilizes only half the value of the DC supply, and
also it is suitable only for motors with even number of phases.
One of the earliest drive circuits introduced for switched reluctance machines
is the C-dump converter, shown in Figure 2.10. In this converter, the stored
magnetic energy is partially diverted to the dump capacitor and recovered by
a single chopper and sent to the DC source.
Figure 2.10 C–Dump Converter.
The C-dump converter has the drawback of disability to provide zero
voltage across the phase winding during commutation period, which results
in increased acoustic noise, and deterioration of the winding insulation.
These drawbacks were overcome by adding a freewheeling transistor, which
may be used for active energy recovery replacing the energy recovery coil
further modifications were proposed to the conventional C-dump converter to
eliminate its drawbacks while maintaining its attractive advantages [42, 43].
The variable DC link converter is an attractive configuration of SRM
drives due to its advantages and its ability of operation in four-quadrant
sensor-less applications.
17
Figure 2.11 Variable DC Link Converter.
The variable DC link could be achieved by inserting a front end
converter which may be a buck or a buck-boost converter. Figure 2.11 shows
a variable DC link converter with a buck-boost front end stage. The machine
input voltage can be varied from zero to twice the supply voltage, further,
faster commutation of the current is achieved. A combination of the C-dump
and the buck front end converters was proposed in [44] as a hybrid converter
that possesses the advantages of both topologies and overcomes their major
drawbacks.
All the above converter topologies are known as hard switching
topologies, since the power switches and diodes are switched ON and OFF
while their voltages and currents are nonzero. The converter topologies that
enable soft switching are known as resonant topologies, one of which is
shown in Figure 2.12 below.
18
Figure 2.12 Resonant Converter.
The attractive advantage of zero switching losses which could be
offered by the resonant circuit is contradicted by the losses in its passive
elements, and the electromagnetic interference, EMI influenced by such
circuits. There are several configurations within the resonant converter
topology, depending on the resonance elements arrangement, (series or
parallel), the soft switching technique (ZCS or ZVS), and the application (is
phase current overlap required or not)[45, 46]. Few researches paid attention
to the resonant converter topology because of the un-encouraging
experimental studies [1].
More detailed study and suggestion of a new configuration based on the
resonance principle and utilizing a switched capacitance circuit to perform as
a variable capacitor is proposed in chapter 5 of this thesis.
The last converter topology to be presented in this section is the two
stage power converter, Figure 2.13 in which the three phase AC supply is
converted to a single phase variable frequency via a controlled
rectifier/inverter, the other stage is the commutating stage which energizes
phase windings [1].
19
Figure 2.13 Two Stage Power Converter.
The two stage converter has the advantage of being capable of
returning the power from the machine to the supply side directly with no
limitations and eliminates the DC link capacitor. A comparison of the above
discussed topologies showing the advantages and disadvantages together
with useful application notes is presented in Table 2-1.
20
Table 2-1 Comparison of Common SRM Drive Circuits
Type Switches
/phase
Diodes
/phase Advantages Disadvantages Application Notes
Asymmetric
Bridge 2 2
-Ideal for high performance
current and torque control.
-Allows greater flexibility in
controlling machine current.
-Capable of +ve, -ve, and
zero voltage output.
-Voltage stress across the
switching element is restricted
to Vdc.
-High switching losses.
-Larger Heat sinks.
-Not suitable for high power
applications.
Low power levels
inverters fed from a
voltage source.
R- Dump 1 1
-Compactness of converter
package.
-Lower cost due to minimum
number of switches and
diodes.
-Unable to apply zero voltage
output during current conduction.
-Reduced system efficiency.
-Rate of change of phase voltage
is doubled.
Low speed, low
switching frequencies.
Bifilar 1 1 -Compactness of converter
package.
-High stress on switching element.
-Reduced power density[1].
Not economical for
large motors.
21
-Lower cost due to minimum
number of switches and
diodes.
-Capability of regeneration of
stored energy.
Split DC
supply 1 1
-Compactness of converter
package.
-Lower cost due to minimum
number of switches and
diodes.
-Capability of regeneration of
stored energy.
-De-rating of the supply voltage.
-Suitable only for motors with an
even number of phases.
Fractional hp motors
with even number of
phases.
C –Dump 1 1
-Minimum switches number.
-Independent phase current
control.
-Current commutation is limited by
the difference between the voltage
across Cd and the DC link.
-Not suitable for high speeds.
-Lower efficiency.
-Unable to provide zero voltage.
Low speed
applications
C –Dump 1 1 -Minimum switches number. -Only motoring operation is Suitable only for
22
with
freewheelin
g transistor
-Independent phase current
control.
-Lower cost than C-dump.
-Higher power density than C-
dump.
-Less ripple on DC link
capacitor.
-Able to provide zero voltage.
possible.
-High rating of the freewheeling
switch.
-complexity of control when phase
currents overlap.
motoring operation.
Minimum
switch with
variable
DC-link
1 1
-Higher DC source can be
adopted
-Lower core losses[1, 44].
-Lower switching losses.
-Reduced acoustic noise.
-Lower commutation voltage.
-Lower overall system efficiency.
-Unsuitable for continuous
generative operation[1].
-Suitable for four
quadrant sensor-less
applications.
-Preferred in
generation mode of
operation.
Resonant
converter 1 1
-Low switching losses
(theoretically zero).
-Superior quality of current
waveform.
-High switching frequency
-Lower power density.
-EMI influence.
-Higher rating of switching
elements.
Suitable for high
frequency applications
(f>20kHz).
23
capability.
Two –stage
power
converter
1(+6) 1(+6)
-Lower cost.
-Higher reliability.
-Higher power density.
-Higher number of switches.
-Not economical if regenerative
operation is not substantial.
Suitable for variable
speed, constant
frequency generation
using wind energy[8].
24
However, the selection of the proper converter for a certain application
depends on many factors the designer has to compromise between them,
such as; the hp capability, size and volume, economic considerations, speed
requirements, switching frequency’s range and whether the machine is used
for motoring or for generation.
From the literature review and the above comparison between different
topologies of converters which are utilised as an SRM drive circuits, it has
been shown that the asymmetric converter topology is the most common
topology which is used for most of the applications of the switched reluctance
machines in industry, therefore this converter will be used after modification
to implement the new converter topology aimed from this research.
25
Chapter 3
Fundamentals and Theory of Operation of SRM
3.1 Motor Construction
The Switched Reluctance Motor (SRM) consists of a salient pole stator
and a salient pole rotor. Only the stator carries windings (wound field coils of
a DC motor), the rotor has neither windings nor magnets and it is built up
from a stack of steel laminations (hence the SRM is referred to as a doubly
salient, singly excited motor) [47].
Figure 3.1 Basic Structure of SRM.
The SRM is classified as an AC synchronous machine, since the
excitation sequence of stator phases should be synchronized with the rotor
position to make the rotor rotates [47].
Each stator phase is consisted by connecting the two diametrically
opposite windings in series, as shown in Figure 3.1.
The switched reluctance motor is similar to the step motor except that
the former has fewer poles, larger step angle, and higher power output
capability.
26
3.2 Theory of Operation
Figure 3.2 shows only two of the stator poles, and two of the rotor poles
to explain how the torque is produced in SRM. At the position shown in
Figure 3.2, the leading edge of the rotor pole meets the edge of the stator
pole (at this position the inductance of phase A is minimum and the
reluctance is maximum). If this stator pole is excited at this moment, (i.e,
both the opposite poles of the stator are energized to constitute opposite
magnetic poles), then the rotor pole is attracted to the stator pole causing the
rotor to rotate in clockwise direction due to the fact that in a magnetic circuit
the rotating member prefers to come to the minimum reluctance position at
the instance of excitation.
As the rotor starts moving, the inductance is increased with the rotor
position until it reaches its maximum value when the stator and rotor poles
are completely aligned (fully overlapped). The inductance remains constant
at its maximum as long as the two poles stay overlapped, this region is called
the maximum inductance dead zone, where the name dead comes from zero
torque production in this period.
Figure 3.2 Single Element of SRM.
θ5
θ1
βs βr
27
As the rotor continues rotating, the poles are not completely overlapped
and the inductance starts to decrease until it reaches its minimum value
when the two poles are apart from each other and it remains at this value
until overlap occurs again, which is known as the minimum inductance dead
zone.
The variation of phase inductance of the stator coil with the angle of
rotation (θ) is illustrated in Figure 3.3.
While two rotor poles are aligned to the two excited stator poles,
another set of poles is out of alignment with respect to a different set of the
stator poles. The principle of operation depends on switching of currents into
stator windings sequentially such that each phase is excited when its
inductance is varying to produce positive or negative torque when the
inductance is increasing or decreasing respectively, and only the sequence
of excitation of stator phases determines the direction in which the rotor will
rotate.
3.3 Torque Production
The Switched Reluctance Motor is an electric motor in which torque is
produced by the tendency of its movable part to move to a position where the
inductance of the excited winding is maximized. During motor operation,
each phase is excited when its inductance is increasing and unexcited when
its inductance is decreasing.
Figure 3.3 Ideal profile of phase inductance.
θ1 θ2 θ3 θ4 θ5
Lu
La
Rotor position
Ph
ase
ind
uct
ance
Rotor pole
28
The air gap is minimum at the aligned position (the position at which a
pair of rotor poles are exactly aligned with stator poles), the magnetic
reluctance to the flux flow is at its lowest; it will be highest at the unaligned
position. Thus, for a given phase, when the rotor is not aligned with the
stator, the rotor will move to align with the excited stator pole [47].
The key to effective control of Switched Reluctance Motor is to switch
the current from one phase to the next and to synchronize phase’s excitation
as a function of rotor position.
The current flowing in each stator phase is defined in magnitude, rise
and fall times by the following parameters: motor resistance (including coil,
wiring and on-state resistance of electronic switches), the inductance of the
motor coil. The rise and fall times are directly proportional to coil inductance
and inversely proportional to resistance.
The relation between the phase inductance, L and rotor position, θ can
be deduced using Figure 3.3. The significant changes in the inductance
profile are determined in terms of the stator and rotor pole arcs and number
of rotor poles neglecting saturation. The stator pole arc is usually less than
the rotor pole arc in most practical cases.
From Figure 3.3, different positions of the rotor can be derived as
follows[1]:
3.1
3.2
3.3
3.4
3.5
Where and are the stator and rotor pole arcs respectively and is
the number of rotor poles.
29
The phase inductance has four distinct regions depending on the
respective rotor position along one complete revolution:
1. Unaligned inductance region: this occurs in the intervals 0-
and - at which the stator and rotor poles are not overlapping
and the flux is predominantly determined by the air path, thus
making the inductance minimum and almost constant. Thus,
these two regions have no contribution to torque production.
2. Poles overlapping region: in this region, - , the stator and
rotor poles begin to overlap and the flux path is mainly through
the lamination of both poles, this increases the inductance with
the rotor position, giving it a positive gradient, hence, if current is
impressed in the winding during this region, it will produce a
positive torque. The inductance reaches its maximum when the
overlap of the two poles is complete.
3. Aligned inductance region: due to the fact that the two pole arcs
of the stator and rotor are not equal, the movement of the rotor
during the region - doesn’t change the complete overlap of
the two poles, hence doesn’t change the dominant flux path, and
the inductance is kept constant at its maximum (aligned) value
giving no contribution to the generated torque. This dead region
serves on the other hand a useful function by allowing more time
for the current to come to a lower value when it is commutated.
4. Poles departing region: in this region, - , the rotor pole
continue moving away from alignment with the stator pole, which
is the opposite process to region 1, hence the inductance change
during this region is similar to that of region 1 but with a negative
gradient.
30
3.4 Equivalent Circuit of a SRM
Neglecting the mutual inductance, an equivalent circuit of the SRM could be
derived as follows:
The instantaneous voltage across the terminals of a single phase of a
SRM winding is related to the flux linked in the winding by Faraday’s law;
equation 3.6
dt
dRiv m
φ+=
3.6
Where, v is the terminal voltage, I is the phase current, Rm is the motor
phase resistance, and φ is the flux linked by the winding, which varies as a
function of rotor position θ, and the motor current I, hence, equation 3.6 can
be expanded as:
dt
d
dt
di
iRiv m
θθ
φφ
∂
∂
∂
∂++=
3.7
Equation 3.7 governs the transfer of electrical energy to the SRM’s
magnetic field. Multiplying both sides of equation 3.6 by I, gives an
expression for the instantaneous power in the SRM, equation 3.8
dt
diRivi m
φ+=
2 3.8
The left hand side of equation 3.8 represents the electrical input power,
while the first term of the right hand side represents the ohmic losses in the
motor winding, thus the second term of the right hand side must represent
the sum of mechanical output power and any power stored in the magnetic
field, since power is to be conserved. This is what equation 3.9 implies.
dt
wd
dt
wd
dt
di
fm+=
φ 3.9
31
Where, Wm is the mechanical energy, and Wf is the magnetic field
energy. We can express the mechanical power as a function of torque, T and
the speed,ω as:
dt
dTT
dt
wd m θω ==
3.10
Substituting from equation 3.10 into equation 3.9 we get:
dt
wd
dt
dT
dt
di
f+=
θφ 3.11
Solving for T from equation 3.11 we get
θθ
φφθ
d
wd
d
diT
f−=),(
3.12
Figure 3.4 Magnetization Curve.
It is often desirable to express torque in terms of current rather than
flux, hence appears the concept of co-energy, Wc instead of energy. To
introduce the co-energy concept, consider the graphical interpretation of field
32
energy, Figure 3.4. Solving equation 3.11 for Wf at a constant shaft angle,
gives:
φφ
diw f ∫=0
3.13
This is represented by the area to the left of the curve of Figure 3.4. The
co-energy for the same fixed angle, θ is given by:
∫=i
c diw0
φ 3.14
Which is represented by the area under the curve of Figure 3.4, it is shown
from figure that the sum of the magnetic field energy and the co-energy for a
fixed angle, θ is equal to the area of the rectangle which area is equal to (φ.i)
Hence we can write;
iww fc⋅=+ φ 3.15
Remember that we want to substitute Wf in equation 3.12 in order to
express the torque in terms of current. Then from equation 3.15 we have:
wddidiwd cf−+= φφ 3.16
Substituting from equation 3.16 into equation 3.12 we get:
θ
φφφφθ
d
wddididiT
c)(
),(−+−
= 3.17
And
dii
wd
wiwd
cc
c ∂∂
∂∂
+= θθ
θ ),( 3.18
For a constant current, from equation 3.17 the torque is simply given by
θd
wdT
c=
3.19
33
It is often assumed in SRM analysis that the motor remains
magnetically unsaturated during operation, when this is assumed the relation
between flux and current is given by:
iL ⋅= )(θφ 3.20
Substituting from equation 3.20 into equation 3.14 we get
)(2
2
θLiw c
= 3.21
And from equation 3.19
θd
dLiT
2
2
= 3.22
Also, substituting from equation 3.20 into equation 3.6 we get
=++=+=⋅
θθθ
d
dL
dt
di
dt
diLRi
dt
iLdRiv mm
])([
θω
d
dLi
dt
diLRi m
++= 3.23
In equation 3.23, the three terms of the right hand side, and by analogy
to a series DC machine, it represent the resistive voltage drop, the inductive
voltage drop, and the induced EMF respectively. Hence, the induced EMF
could be given as
Kid
dLie bω
θω ==
3.24
Where Kb may be defined as the EMF constant similar to the DC series
excited machine and is given here as dL/dθ, which means that it depends on
the rotor position and is obtained for a constant current at that position.
From the above derivation, the equivalent circuit of an SRM phase could be
derived as shown in Figure 3.5.
34
Figure 3.5 Single phase equivalent circuit of a SRM.
3.5 Static Characteristics of the SRM
The static characteristics of SRM include the flux linkage current
curves, co-energy curves, and the static torque curves. In the following
section, the experimental measurements of the parameters of the SRM is
shown as stated in [48], which is the same parameters used in this case
study.
3.5.1 Flux Linkage Current Curves
Flux linkage current curves represent the relationship between the
phase flux linkage and phase current at different rotor positions. The two
curves at the extreme aligned and unaligned rotor positions are considered
experimentally, and the intermediate curves are obtained analytically.
3.5.1.1 Aligned rotor position
The flux linkage current curve at aligned rotor position is obtained
experimentally as follows:
1- A certain phase of the motor is supplied by DC supply to ensure that
the rotor is settled exactly at the aligned position. By measuring the
current and DC supply, the phase resistance is obtained.
2- Then, the test is performed by exciting the same phase windings by a
variable AC voltage. The current is measured at different voltage
levels, and these data are used to obtain the flux linkage as a function
of the excitation current.
3- If the phase current iac is measured at a certain voltage Vac, then the
phase impedance Zph is given by:
35
ac
ac
phi
VZ = 3.25
And the phase reactance Xph can be calculated from:
22
phph RZX −= 3.26
Where R is the phase resistance (is known from DC test).
4- The phase inductance is calculated by:
f
XL
ph
π2= 3.27
Where f is the frequency of AC supply (50 Hz.).
5- The flux linkage φ corresponding to the measured current is
obtained by:
iL ⋅=φ 3.28
This procedure is repeated for different levels of excitation current to
obtain the flux linkage as a function of that current.
3.5.1.2 Unaligned rotor position
The rotor must be fixed such that the poles of the excited phase are
facing an inter-polar space of the rotor, since the rotor is at unaligned
position. To achieve this position, two aligned positions are marked as
shown in the previous test, for two different phases. The rotor is hold on
mechanically in the midway between these two marked positions. At this
position, one of these two phases is excited with AC voltage source, and
the current is measured.
The flux linkage is obtained by using the same procedure discussed,
from step 3 to step 5, for last subsection. There is no possibility for
magnetic saturation at the unaligned position under normal operating
condition. So, one value of flux is enough to get the slope of this
characteristic, but for more accuracy more points are calculated.
Figure 3.6 shows the measured flux linkage current curves at both the
two extreme aligned and unaligned rotor positions.
36
0 2 4 6 8 10 12 0
0.5
1
1.5
2
Current (Amp.)
Flu
x l
ink
age
(Web
.)
Aligned
Unaligned
Figure 3.6 Measured flux linkage-current curves at aligned and unaligned positions.
3.5.1.3 Flux linkage current curves at intermediate rotor positions
As a complete view of flux current characteristics, the flux current
curves at other intermediate rotor position angles must be obtained
analytically after obtaining the measured flux linkage curves at both
unaligned and aligned positions.
An analytical function that represents flux linkage characteristics over
the entire half rotor pole pitch may be expressed as [49]:
)()(),(0
θφθφ rn
N
n
nPCosii ∑=
= 3.29
Where φn(i) are suitable expression depending only on the current
variable, Pr is the number of rotor poles.
The first two terms in equation 3.29 are considered, because there are
only two known available values for both aligned and unaligned rotor
positions.
)()()(),( 10 θφφθφ rPCosiii += 3.30
37
To find the values of φ0(i) and φ1(i) consider the following procedure:
By substituting in equation 3.30 assuming unaligned position (θ = θu =
0), also at aligned position (θ = θal = 2π / Pr). Then;
)()()( 10 iiiu φφφ += 3.31
)()()( 10 iiia φφφ −= 3.32
Solving the set of equations (3.31), and (3.32), then;
)]()([2
1)(0 iii au φφφ += , and 3.33
)]()([2
1)(1 iii au φφφ −= 3.34
Where φu, and φa are the flux linkage at both unaligned and aligned
positions.
By substituting equations 3.33, and equation 3.34 in equation 3.29, an
expression for the flux linkage characteristics is obtained:
)](1[2
)(),( θ
φθφ r
ua
u PCosiLi
iLi −−
+= 3.35
By choosing a different values of θ (between θ = 0, and θ = π / Pr), the
intermediate (φ, i) curves are obtained using equation 3.35. Usually choosing
the step with equal spaces rotor position angles. In the current case, the
intermediate flux linkage current curves obtained by spacing three
mechanical degrees. The number of (φ, i) curves between aligned and
unaligned positions are equal to fourteen curves as shown in Figure 3.7.
38
0 2 4 6 8 10 12 14 0
0.5
1
1.5
2
Current (Amp.)
θ =0o
θ =12o
θ =24o
θ =36o θ =45
o
Flu
x l
ink
age
(Web
.)
Figure 3.7 Flux linkage-current curves of 6/4 SRM.
Current (Amp.) Position (Deg.)
Flu
x l
ink
age
(Web
.)
Figure 3.8 The three dimension of flux linkage, current, and position angle curves.
39
Thus, the flux linkage current data along half rotor pole pitch is
completed, this data will be used to compute the co-energy curves, and it is
reversed and stored in the form of i(θ, φ) to be used later when solving the
motor differential equations. Figure 3.8 illustrates the three dimensions of flux
linkage, current, and rotor position curves.
Therefore, the vector of flux linkage can be obtained from the last
estimated curves for any values of rotor position angle and stator current or
by interpolation if it lies between any two closest curves as shown in Figure
3.8. The current values can be obtained form the corresponding values of
both flux linkage and rotor position angle. Then, all data are stored in a look-
up table matrix i(θ, φ). In this matrix, the current values are arranged such
that the flux linkage values represent the row index and rotor position angles
represent the column index [48].
3.6 Torque-Speed Characteristics
The torque-speed operating point of the SRM is essentially
programmable, and determined almost entirely by the control. This is one of
the features that makes the SRM an attractive solution. The envelope of
operating possibilities is limited by physical constraints such as the supply
voltage and the allowable temperature rise of the motor under increasing
load. In general this envelope is described by Figure 3.9 [2].
Figure 3.9 SRM Torque-Speed Characteristics.
40
Like other motors, torque is limited by the maximum allowed current,
and speed is limited by the available bus voltage. With increasing shaft
speed, a current limit region persists until the rotor reaches a speed where
the back EMF of the motor is such that, given the DC bus voltage limitation
we can get no more current in the winding, thus no more torque from the
motor. At this point, called the base speed, and beyond, the shaft output
power remains constant at its maximum. At higher speeds, the back EMF
increases and the shaft output power begins to drop. This region is
characterized by the product of torque and the square of speed remains
constant [2].
3.7 Control Strategies of Switched Reluctance Motor Drives
There are different well known strategies of machine control depending
on the application, these are, position, speed, current, and torque control.
The switched reluctance machine is different from other types of DC and AC
machines because of its poles saliency and nonlinearity. Further more, the
current control strategy in SRM is different from torque control while these
two are synonyms in DC drives.
The SRM produces discrete pulses of torque, and the proper design of
the machine which allows overlapping inductance profile could make it
possible to produce continuous torque. The developed torque is controlled by
adjusting the magnitude of the phase current or by varying the dwell angle,
which is recommended to be kept constant in order to reduce the torque
ripples [1]. The control operation may be done using a position sensor
feedback signal, or using sensor-less operation, by estimating the position
from the magnetic characteristics of the machine [50-55].
In the following sections, a brief overview of the control principle of
different schemes is presented.
41
3.7.1 Speed Control
The speed control scheme, Figure 3.10, consists of two loops, current
loop and speed loop through which the speed error is processed and
conditioned with the speed controller to obtain the current command signal,
i*, this current command is compared to the actual phase current feedback
signal to generate the current error signal. This later is processed through
the current controller to produce the command signal for the converter, v*,
and the switching of the phase is determined according to a predetermined
window, ∆i, of the hysteresis current controller. The currents are diverted into
each phase according to its position information which maybe measured or
estimated. In general, the speed control strategy consists of two actions, the
first is to regulate the motor speed by adjusting the duty cycle of a PWM
signal that activates the power converter, and the second action is to adjust
the advanced firing angle as a function of the motor speed to improve motor
efficiency and performance specially at high speeds [56].
Figure 3.10 Speed Control Closed-loop Block Diagram.
3.7.2 Current Control
The SRM is operated either in current control or voltage control mode.
The voltage control operation mode is cheaper in implementation but is not
convenient for servo-type applications, in this case current control operation
mode is essential [57].
1
Speed
Speed Controller MotorCurren t Contro ller Converte r
1
Reference Speed1i* v * v
Current
Speed
Speed
42
The most significant parameter in the current control loop is the
hysteresis band, which is desired to be as small as possible in order to
reduce current ripples and to increase the bandwidth of the current loop,
however the residual current ripple content, the high-variable switching
frequency are the un-encouraging drawbacks of this approach [57].
3.7.3 Torque Control
Unlike DC and AC machines, the air-gap torque in the SRM has a
nonlinear relationship with the excitation current which makes the control
system more complex, since the torque is depending not only on the
excitation current but on the rotor position as well. The method of torque
control depends on the number of phases which are excited at the same
instant, which severely affects the torque ripple content and consequently the
speed ripples which in turn reduce the overall efficiency and performance of
the drive system [1].
The designer of a SRM drive system has to aim at one of two conflicting
concerns, maximizing the output torque, or maximizing the efficiency. The
efficiency is maximized by minimizing the dwell (conduction) angle of the
phase, while maximum torque is obtained when the dwell angle is
maximized. In general, two strategies of dwell angle are followed, either
constant or variable dwell angle, depending on the application. Different
modes of SRM commutation are given in [40].
Although the switched reluctance machine has a strong similarity to the
series excited DC and synchronous machines, it is different from these
machines from the control point of view, more complications arise in the case
of SRM due to the fact that the variable inductance of the motor is not only a
function of the rotor position, but the phase excitation current as well, in
contrast, for other machines, control strategies are based on the machine
parameters being almost constant for the control range.
The control requirement could be classified to low performance and
high performance categories based on torque ripples, and speed response
specification.
43
The heart of any motor drive’s control system is current control [1]. In
most of the DC machine drives, torque control is synonymous with current
control, but because of the nonlinearity of the SRM, this concept is
significantly different.
3.8 Control Principle of SRM
Referring to Figure 3.11, the phase winding is switched ON during the
positive gradient period of the inductance profile for motoring operation. The
SRM produces discrete pulses of torque, and it could produce a continuous
torque if proper design of overlapping inductance profile of different phases
is achieved, on the other hand this overlap reduces the power density of the
motor and complicates the control of the SRM drive [1].
Figure 3.11 Switching strategy for motor operation.
44
The average torque is controlled by adjusting the magnitude of the
phase current or by varying the dwell angle, θd. It is recommended, to reduce
torque ripples, to keep the dwell angle constant and vary the magnitude of
the phase current, which requires a current controller to be within the drive
circuit of the motor adding a safe operation feature to this approach.
It is essential to divert the current into the winding at the instant of
increasing inductance to ensure instantaneous torque production, but due to
the fact that the current can’t rise or fall instantaneously in an RL circuit, the
voltage has to be applied with a certain advance angle, θa , Figure 3.11, and
for the same reason current has to be commutated at a certain advance
commutation angle θco in order to bring the current to zero before the
negative gradient of the inductance profile is reached so that no negative
torque may be produced. These two advance angles are function of the
magnitude of the phase current and the speed of the rotor.
Figure 3.12 shows typical current waveforms for various advance
angles and dwell angles [1].
45
Figure 3.12 Typical current waveforms for various advance and dwell angles.
The shaft position feedback information is essential in all control
strategies for generating switching commands for the power converter for the
position feedback loop, and for the speed feedback loop which can be
derived from the position data. There are several types of position sensors
commonly used which are;
- Photo-transistors and photo-diodes.
- Hall effect elements.
- Pulse encoders.
- Variable Differential Transformers.
In the current control technique, the magnitude of the current flowing
into windings is controlled using a control loop with a current feedback. The
current in a motor phase winding is directly measured with a current/voltage
46
converter or a current sense resistor connected in series with the phase and
compared with the desired value forming an error signal. This error is
compensated via a control scheme such as PID and an appropriate control
action is taken [47].
This method has the disadvantage that even if the phase current is
maintained constant, the torque produced is not smooth not only because of
control strategy and phase commutation but also because of the nonlinear
relationship between torque and current. The torque ripples are undesirable
because it contribute to the problem of audible noise, vibrations, and also
introduce torque disturbances which manifest as velocity errors.
A solution to the problem of torque ripple is to profile the current such
that torque is the controlled variable (torque control). Since torque can’t be
sensed directly, this can only be done using a prior information about the
motor’s torque-current-angle characteristics. The torque control strategy is
based on following a contour for each phase of the SRM such that the sum
of torque produced by each phase is constant and equal to desired torque,
this desired total torque is calculated from the speed loop and is split into
desired phase torque via shaping function [47].
In the following chapter, a new technique utilising the switched
capacitance circuit to profile the phase current is discussed in details.
3.9 Sensor-less Operation of SRM Drives
Sensor-less operation of the SRM is desirable because of compactness
in weight and volume, also lower cost is achieved due to the elimination of
the mechanical assembly and mounting of the rotor position sensors. The
driving factor behind sensor-less operation of any machine drive system
including that of the SRM is the quest for low cost with high performance
drive. Three broad classes of rotor position estimation have been emerged in
research, these are the observer based scheme, the incremental inductance
based measurement using current magnitude or fall and rise time and finally
the inductance based estimation. A brief description of each of the previous
techniques is given in the following sections [1].
47
3.9.1 Incremental Inductance Measurement [1]
This method uses the fact that the current rise and fall times are
proportional to the incremental inductance, and under certain assumptions
the rise and fall times reflect the incremental inductance and hence the rotor
position itself. The only advantage of this method is that it only needs to
measure the current and its rise and fall times to predict the rotor position
and can use the active phase itself while it has many disadvantages.
3.9.2 Observer based rotor position estimation [1]
This technique depends on the real time solution of the SRM system
equations which gives estimates of the current, speed and rotor position. To
obtain this, it is required to know the rotor speed dynamics obtained through
the electromechanical equation coupling the air gap torque to the mechanical
system with rotor inertia and friction constants. For estimation, the machine
model incorporating the error variables with their gain adjustment feature
constitutes what is known as an observer.
3.9.3 Inductance Sensing
As the direct rotor position sensors don’t provide any information about
electrical characteristics of the motor because they are insensitive to
inductance profile variation with rotor angle, and as the torque production is
actually dependent on the rate of change of co-energy with rotor position, it
appears that a desired instantaneous torque can be obtained from
instantaneous inductance information rather than rotor position [47]. Since
there is at least one idle phase in a SRM, the inductance of that phase can
be sensed for the purpose of commutation control. The inductance can be
estimated from the measurements of the motor terminal voltages and
currents.
48
3.9.3.1 Phase Pulsing
A voltage pulse V is applied to an un-energized phase by the drive
converter for a period of time ∆T and the change in coil current ∆I is
measured. The inductance is obtained from:
I
TL
∆∆
= 3.36
3.9.3.2 Frequency Modulation
Inductance information is encoded in a frequency modulated signal
using a low voltage analogue circuit.
3.9.3.3 Phase Modulation
A low alternating voltage is applied to an un-energized phase and the
phase angle difference between the input voltage and the resulting current is
detected, the inductance is given by:
ω
ϕtanRL =
3.37
Where, ϕ is the phase angle.
3.9.3.4 Amplitude Modulation
A low level alternating voltage is applied to an un-energized phase and
the amplitude of the resulting current is mapped to the coil inductance. The
inductance can be expressed as:
RI
VL
m
m 2
2
2
1−=
ω
3.38
Where, Vm is the amplitude of the input alternating voltage, Im is the
current amplitude, and R is the resistance in the circuit.
49
3.9.3.5 Self-Voltage Technique
The inductance of the active phase is estimated in real time from
measurements of the active phase current and phase flux. If Iph is the current
in the active phase linking a flux φph then the phase inductance is given by:
IL
ph
ph
ph
φ=
3.39
3.9.3.6 Artificial-Intelligence-based inductance estimation [58]
Most of the difficulty in understanding the operation and design of the
switched reluctance machine drives originates from the fact that the switched
reluctance machine is a doubly-salient, highly nonlinear machine. This
makes it an ideal candidate to incorporate artificial intelligence in a switched
reluctance machine drive.
Recently, extensive research work has been conducted in the field of
sensor-less operation of switched reluctance machines utilising artificial-
intelligence-based inductance estimation, artificial-intelligence-based position
estimation and artificial-intelligence-based electromagnetic torque estimation
as well as using artificial intelligence techniques for machine design and
control. A brief illustration of some artificial-intelligence-based techniques is
presented in the following sections.
As discussed earlier, the switched reluctance machine has a phase
inductance which is a nonlinear function of the stator phase current and the
rotor position. It is possible to use a fuzzy-logic-based estimator to estimate
the phase inductance of the SRM. The inputs to this fuzzy estimator are the
stator phase current and the rotor position while the output is the phase
inductance. The universe of discourse for the two input variables is divided
into an appropriate number of fuzzy sets for the stator current and an
appropriate number of fuzzy sets for rotor position. In the fuzzification stage,
the corresponding membership function values are obtained respectively.
For this purpose, 7 triangular membership functions are used for the rotor
50
position, where the corresponding fuzzy sets are B1 to B7, shown in Figure
3.13 and 9 triangular membership functions for the stator current, where the
corresponding fuzzy sets of which are A1 to A9. The universe of discourse for
the output variable which is the phase inductance in this case is divided into
11 fuzzy sets, C1 to C11 in Figure 3.13, in which, C1 is used for the smallest
inductance value which physically corresponds to zero rotor position and
smallest stator current, while for the largest inductance value C11 is used.
Figure 3.13 Input and output membership functions
for fuzzy inductance estimation.
The phase inductance is determined by using the appropriate If-Then
rules of the fuzzy-logic estimator which in general take the form “if i is A and
θr is B then L is C”, where i, θr and L are the stator phase current, the rotor
position and the phase inductance respectively.
The fuzzy output is computed by using the sup-min method, while the
crisp output can be calculated in the defuzzification stage by using the height
defuzzification technique described in [58], the crisp output in this case is the
actual phase inductance.
Figure 3.14 shows an example of the estimated phase inductance being
compared with the experimentally obtained one.[58]
51
Figure 3.14 Phase inductance, (a) Experimentally
obtained. (b) fuzzy-logic-based.
3.9.3.7 Artificial-Intelligence-based position estimation
Besides the conventional techniques used to estimate the rotor position
in a switched reluctance machine, the universal function AI-based position
estimators are ideal candidates in position-sensor-less SRM drives.
Figure 3.15 shows an ANN-based position estimator where it is
assumed that the flux linkage-current characteristics of the switched
reluctance machine are known and the ANN is used as a nonlinear function
approximator of these characteristics.
The ANN can be a multilayer feed-forward neural network and the
training can be based on back-propagation. [58]
The input to train the ANN are the input data pairs of the flux linkage, φs
and the stator current, is while the corresponding output is the rotor position,
θr.
52
Figure 3.15 ANN-based position estimator.
However, to decide which side of the alignment the rotor is, angles from
more than one phase has to be estimated. In the shown system of Figure
3.15, the stator voltage, us and the stator current, is are measured, and the
flux linkage, φs is calculated from the measured values of us and is by
integration. In this technique, errors are expected to occur which leads to an
inaccurate estimation of the rotor position especially near the two extreme
positions, the aligned and the unaligned position.
It is possible to have more accurate ANN-based position estimator
where the role of the neural position estimator is not to approximate the static
characteristics, but its role is to approximate the relationship between the
output quantity which is the rotor position and the inputs which are the stator
current and the stator flux under real operating conditions. In this case, an
initial test stage where the motor is run under its real operating conditions
and the input and output data are measured and used to train the ANN, a
position sensor is needed only during this training data collection stage.
It is also possible to have another ANN-based position estimator
scheme where the ANN directly estimates the position by using the
monitored stator voltage and current. The training of this ANN can be used
on data which is obtained in an initial test phase where apposition sensor is
needed as well. Also, a fuzzy-neural estimator of Mamdani or Sugeno type
can be used to replace the ANN to eliminate the problem of finding the
number of hidden layers and hidden neurons.
53
Another system which uses an adaptive network fuzzy inference system
(ANFIS) to estimate the rotor position was illustrated in [10] to estimate the
rotor position of a 6/4 switched reluctance motor. This system which is
shown in Figure 3.16 uses a first order Sugeno-style fuzzy system with two
inputs, which are the flux linkage and the stator currents, and one output,
which is the rotor position.
Figure 3.16 ANFIS structure for 2 inputs and 1 output.
An indirect method for rotor position estimation both at standstill and
during running conditions has been presented in [10] by applying a reduced
voltage to the stator phase windings for a short duration, the initial rotor
position is accurately predicted, which ensures the starting of SRM at any
initial rotor position without starting hesitation and also in the desired
direction of rotation. The proposed method gave an accurate estimation of
the rotor position with an error less than 0.1% for low current while the error
was less than 0.5% for high current. This position estimation algorithm with
the shown error percentages, ensures a sensor-less high performance and
reliable SRM drive operation.
54
3.9.3.8 Artificial-Intelligence-based electromagnetic torque estimation
AI-based techniques could be used to estimate the electromagnetic
torque in a switched reluctance machine drive, for this purpose, neural or
fuzzy-neural networks can be used. [58]
The electromagnetic torque produced by a switched reluctance machine
is a function of the rotor position and the stator flux linkage which in turn is a
nonlinear function of the stator current. This is why it is difficult to implement
a mathematical-model-based estimator of the electromagnetic torque, for this
purpose an AI-based estimator could be used, where the inputs to this
network can be the stator current and the rotor position.
To increase the robustness of the estimator, extra inputs can also be
used such as, the turn-off angle and the stator current limit. Suggestion of
the type and structure of such a network is given in [58].
Another ANFIS model is presented in [59] where an indirect method to
measure the static flux linkage was used and then the co-energy method (via
the principle of virtual displacement) was used to calculate the torque
characteristics from data on flux linkage versus current and rotor position. A
hybrid learning algorithm, which combines the back propagation algorithm
and the linear least-squares estimation algorithm, identifies the parameters
of the ANFIS. After training, the ANFIS flux linkage model and ANFIS torque
model become in agreement with experimental flux linkage measurements
and the calculated torque data. In the same research, an ANFIS current
model and an ANFIS torque model were used to study SRM dynamic
performance. The structure of the ANFIS and the ANFIS models for the
estimation of flux linkage and torque are shown in Figure 3.17 and Figure
3.18 respectively.
55
Figure 3.17 Structure of the ANFIS.
Figure 3.18 ANFIS models
(a) for calculating flux linkage (b) for calculating torque.
3.9.3.9 Artificial-Intelligence-based SRM design
The AI-based technique can be used to obtain the geometric
parameters of a switched reluctance motor to meet a certain operating
condition such as minimizing the torque oscillations at low-speeds.
An ANN is presented for this purpose which is a multilayer feed forward,
back-propagation ANN, whose inputs are the geometrical parameters, the
magneto-motive force and the rotor position, while its output is force.
56
The torque ripples in a switched reluctance motor can be minimised
either by designing the motor with an optimum geometry or by shaping the
stator current profile. It is useful, even if the stator current is shaped, it is
useful to design the doubly salient geometry that minimizes the torque
ripples. In Figure 3.19, an ANN-based force estimator is shown, the input
layer of this ANN has 5 neurons and the five inputs are the three geometrical
parameters (where Λ is the tooth pitch, δ is the air gap width, ts is the stator
tooth width and tr is the rotor tooth width) and the other two inputs are the
normalised rotor position, θrn=θr/Λ and the magneto-motive force, F.
Figure 3.19 ANN-based force estimator.
The number of nodes in the hidden layers is found by trial and error.
More details of the AI-based control of the switched reluctance machine is
given in [58].
Another AI-based system is presented in [60], where a Genetic-Fuzzy
Algorithm (GFA) is used to optimise the design parameters of a 8/6 four
phase switched reluctance machine with two objective functions, the
efficiency and the torque ripples.
Figure 3.20 shows a flow chart of that algorithm while Table 3-1 shows
a comparison between the different design parameters and objective
functions obtained by the initial design, the fuzzy method and the GFA
method. Where, the shown parameters in the table represent; the stator
pole-arc, the rotor pole-arc, the rotor tooth-height, the rotor internal radius,
the stator tooth height, the stator steel thickness, the stator external
diameter, the rotor external diameter, the number of turns per stator-phase,
the efficiency and the torque ripple percentage respectively.
57
Figure 3.20 Optimal design of SRM by GFA.
Table 3-1 Comparison of design parameters and objective functions.[60]
Parameter βs° βr° hr
(mm)
yr
(mm)
hs
(mm)
ys
(mm)
Ds
(mm)
Dr
(mm)
Nph η% TR%
Initial 21 23 19.4 12.9 25.8 17.4 183.5 96.5 28 92.7 28.44
Fuzzy 20 23 19.3 15 23 15 176.9 100.3 28 92.1 15.4
GFA 20.3 23.7 19.2 14.6 20.5 14.7 170.3 99.3 26 93.1 10.4
The results as stated in [60] have shown an increase of 0.4% in the
efficiency and a reduction of about 18% in the torque ripples compared with
the initial design parameters.
The use of artificial intelligence have become familiar in the field of
electrical machine drives in general and in particular in the field of switched
reluctance machine drives as it is a very useful tool in the rotor position
estimation which allows sensor-less operation of the switched reluctance
machine, besides using it as has been shown in previous sections in
58
inductance estimation, torque estimation and the geometrical design of the
motor.
As a matter of fact, the artificial intelligence is very useful to implement
the designed system in this research as it will help to estimate the accurate
values of the phase inductance at each switching instant of the switching-
capacitance, as it is very critical in this case study to find the accurate value
of inductance to resonate with the switched capacitance, as well as
calculating the duty cycle, also the artificial intelligence techniques will be of
great help with synchronising the switching instants of the main switches of
the phase with the switches of the switched-capacitance circuit that is
besides the improvement in the overall design and performance of the
machine drive system as shown above.
59
Chapter 4
Switched Capacitance circuit applied to an SRM Driver topology
4.1 Proposed technique to profile the phase current
The idea of this technique aims at producing a current profile which is
close to the ideal current waveform (ideally a square pulse) during the torque
production period, which means, a current waveform with fast rising time,
ideally zero, a flat top to produce maximum torque, a fast fall time, ideally
zero. The ideal case is not achievable in practice, but the more close to ideal,
the higher overall efficiency and power to weight capacity we get.
In the conventional converter systems utilized with SRM, switching on
the current is done at a certain advance angle in order to allow the current to
build up before the phase inductance reaches the positive gradient region,
which leads to excessive conduction losses with zero torque production for a
considerable period. Similarly, switching off the current is done at a certain
advance angle in order to avoid the negative gradient region of the
inductance, which leads to considerable reduction of the developed torque,
hence the importance of profiling the current in the previous way is
appreciated.
4.2 Switched capacitance circuit
There are a considerable variety of switched capacitance circuits
depending on the number of capacitors and the number of switches utilized.
Only two types of switched capacitance circuits will be considered here,
these are the single capacitance double switch, SCDS and the double
capacitance double switch, DCDS circuits for efficiency and size
considerations of the drive circuit, as using a DCDS circuit with a 3-phase
motor for example, means adding 6 capacitors and six semiconductor
60
switches to the drive circuit, which means, increased volume and weight and
more switching losses.
4.2.1 Single Capacitance, Double Switch (SCDS) Circuit
A schematic diagram of a single capacitor double switch circuit is shown
in Figure 4.1. The circuit consists of a capacitor, C and two semiconductor
switches, S1 and S2 connected as shown.
Figure 4.1 Single Capacitance Double Switch Circuit.
The switches S1 and S2 are switched ON and OFF alternately, i.e. if S1
is ON, S2 is OFF and vice versa. The effective capacitance, Ceff is the
capacitance measured from terminals XX, which will be proven to vary with
varying the duty cycle of the switching element resulting in a variable
capacitance within the range between C and infinity for duty cycles of 1 and
zero respectively.
To deduce the mathematical equations which describe the circuit
operation, a current limiter should be connected to the circuit to limit the
transient current value through the switches, this current limiter is an inductor
and a small resistor connected in series with the switching capacitance
circuit, as shown in Figure 4.2, the only difference which should be made
when applying this circuit to an SRM motor phase, is to replace the fixed
inductor with a variable inductor which varies as a function of the rotor
position.
61
Figure 4.2 Switched capacitance circuit with a current limiter.
The circuit shown in Figure 4.2 has two branches connected alternately
to the supply, the first branch contains L, R and C connected in series when
S1 is ON, and the second branch contains L and R connected in series when
S2 is ON.
The voltage equations of these two branches are as follows, depending
on the duty cycle, D of the switching element, S1:
At D=0;
dt
diLRtitv s
+= )()( 4.1
At D=1;
)()()( tv
dt
diLRtitv cs
++= 4.2
And the instantaneous current, i(t) is given by:
dt
tvdcti
c )()(
)=
4.3
If the supply voltage is given by;
)cos()( ψω += tVtv ms
4.4
The instantaneous value of the current for each case is found to be:
For D=0;
)cos()( 111 ϕω −= tIti m
4.5
Where;
R
Lωϕ tan
1
1
−=
4.6
And
62
)(22
1
LR
VI
m
m
ω+
= 4.7
For D=1;
)cos()( 222 ϕω −= tIti m
4.8
Where;
R
cL )(
tan
1
1
2
ωω
ϕ−
=−
4.9
And
)1
(
2
2
2
cLR
VI
m
m
ωω −+
=
4.10
The steady state capacitor voltage, Vc will be given by equation 4.11
)sin(2)( ϕωω
−= tc
IV
m
ct
4.11
if we adjust R so that R << ωL and R<< cω
1, the phase angles ϕ1 and ϕ2 will
be ± (π/2).
4.2.1.1 The Switching Function
Switching function is defined as the closing and opening interval pattern
of switching by considering equal to 1 whenever the switch is ON and zero
whenever the switch is OFF.
Figure 4.3 Switching Function.
63
The switching function takes the form of a train of pulses of a unity
amplitude, Figure 4.3. The complexity of the generation of this train of pulses
arises due to the fact that it has to be pulse width modulated in a pattern that
controls the value of the equivalent capacitance between points XX, Figure
4.1. In the following is the derivation of the relation between the duty cycle of
switch S1, and the effective capacitance between XX.
Let the switching frequency be fs, which means a period, T and an angular
frequency ωs, where;
fT
s
1=
4.12
And
fss πω 2= 4.13
The angular duration of the unit-value period is 2πD rad, where D is the
duty cycle, and the boundaries of the unit value period with respect to the
time zero are 0 and 2πD rad.
From above the switching function can be expressed as:
∑∞
=
+=1
)sin()cos()(n
nn tnBtnAtf ss ωω 4.14
Calculating for the coefficients of Fourier expansion, we get:
∫+
===tt
t
on
DTtdttf
TA
on
1
1
)(1
0
4.15
== ∫+ tt
ts
on
dttntfT
A n
1
1
)cos()(2
ω
[ ] [ ]
−+= tnttn
nT sons
s
11 sin)(sin2
ωωω
4.16
== ∫+ tt
ts
on
dttntfT
B n
1
1
)sin()(2
ω
[ ] [ ]
−+−= tnttn
nT sons
s
11 cos)(cos2
ωωω
4.17
Substituting from equations 4.15, 4.16 and 4.17 into equation 4.14, we get:
64
( [ ( ) ] ]ttnnT
Dtfonsn
s
+∑+= ∞
= 11 sin2
)( 1ω
ω
[ ] ) [ tntn ss ωω cossin 1⋅−
( [ ( ) ] [ ] ) [ ] tntnttn ssons ωωω sincoscos 11⋅−+−
4.18
For simplification, the switching instant is selected to be t1=-(ton/2) which will
reduce equation 4.18 to:
[ ]tnnT
tn
Dtf sn s
ons
ωω
ω
cos
sin4)(
1
2⋅
+= ∑∞
=
4.19
And by substituting ωs=2π/T into equation 4.19, we get
( )( )tn
n
nDDtf s
nω
ππ
cossin2
)(1
⋅+= ∑∞
=
4.20
In the preceding calculation, )(tf is the switching function of switch S1,
and the switching function of switch S2 will be the complement function of it,
hence
( )( )tn
n
nDDf s
n
t ωπ
πcos
sin21
1
'
)( ⋅= ∑−−∞
=
4.21
The voltage across points XX, Figure 4.1, is given by:
)()( tvtfv cXX⋅=
4.22
Where;
∫ ⋅=t
c dttitfc
tv0
)()()(1 4.23
The voltage across the inductor,
Figure 4.2, is given by:
[ ] [ ]=⋅+⋅= )()()()()( ' titfdt
dLtitf
dt
dLtv L
[ ])()()()( ' titftitfdtd
L ⋅+⋅= 4.24
Substituting f(t)+f’(t)=1, then:
65
)()( tidt
dLtv L
⋅= 4.25
Similarly the voltage across the resistor is given by:
)()( tiRtv R⋅=
4.26
Hence we can rewrite the voltage equation of the circuit,
Figure 4.2 as follows:
dttitftfc
tidt
dLtiRtv
t
s )()()()()()(0
1⋅++⋅= ∫
4.27
If the value of R is selected so that R << Xc and R<<XL, then the resistive
term in equation 4.27 could be neglected, therefore it could be rewritten as:
dttitftfc
tidt
dLtv
t
s )()()()()(0
1⋅+= ∫
4.28
According to [61], if the switching frequency is high ( 10≥m ), the above
equation could be simplified for only the fundamental components to:
[ ] )()(2
tiXXDtv Lcs⋅−= 4.29
Where Xc and XL are calculated at the supply frequency.
The peak value of the voltage vxx is given in terms of total branch reactance
by:
IXDv cXX
^2
^
= 4.30
And the effective capacitive reactance is given by:
I
vX
XX
c eff^
^
= 4.31
Substituting from equation 4.31 into equation 4.30 we get:
XDX cc eff
2=
4.32
Hence;
D
cc eff 2
= 4.33
66
Figure 4.4 Relation between D and Ceff for SCDS circuit.
It is clear from the above discussion that this circuit could be used as a
variable capacitor, depending on the duty cycle, starting with the value of the
fixed capacitor at D=1, and increasing gradually with the reduction of D in an
exponential proportion, Figure 4.4.
4.2.2 Double Capacitance, Double Switch (DCDS) Circuit
Another topology of the switched capacitance circuit, is the Double
capacitance Double Switch circuit, Figure 4.5. In this topology, there are two
branches of a fixed capacitor in series with a semiconductor switch in each
branch.
The principle of operation of the DCDS circuit is the same as that of the
SCDS one, hence the derivation of the relation between the duty cycle and
the effective capacitance between A and B is similar, with the difference that
it might be referred to either of the two branches C1 or C2.
67
Figure 4.5 Double Capacitance Double Switch Circuit.
Using the same approach which has been used in the SCDS case, and
applying it to Figure 4.6, we can write the voltage equations as stated in
equation 4.34.
Referring to branch C2, and at D=0, voltage equation is given by
)()()( 1 tvdt
diLRtitv cs
++= 4.34
While at D=1, the voltage is given by equation 4.35
)()()( 2 tvdt
diLRtitv cs
++= 4.35
Figure 4.6 DCDS topology with a current limiter.
The current, i(t) is given by equations 4.36 and 4.37 at D=0 and D=1
respectively.
68
dt
tvdcti
c )()(
1=
4.36
dt
tvdcti
c )()(
2=
4.37
The voltage across AB is given by equations 4.38 and 4.39 as referred
to branch C1 or branch C2 respectively. [61]
IXDXDv ccAB
^
2
2
1
2^
][ )1( ⋅+= −
4.38
IXDXDv ccAB
^
1
2
2
2^
][ )1( ⋅+= − 4.39
From equation 4.31 and referring to branch C1 we get
][2
2
1
2
^
^
)1( XDXDI
vX cc
AB
c eff
−+==
4.40
Where Xceff is the effective capacitive reactance between A and B
cD
cD
c eff 2
2
1
2 111)1(
ωωω⋅+⋅= −
4.41
From which;
cD
cD
c eff 2
2
1
2 111)1( −+=
4.42
cccDcD
c eff 21
1
2
2
2
)1(1 −+=
4.43
Taking the reciprocal of equation 4.43 gives the value of Ceff in terms of
the fixed capacitors of the two branches and the duty cycle referred to
branch C1.
cDcD
ccc eff
1
2
2
2
21
)1( −+=
4.44
69
Dividing both the numerator and the denominator of equation 4.44 by C1
we get:
)1(2
1
22
2
Dcc
D
cc eff
−+⋅
=
4.45
Let;
cc
1
2=γ 4.46
Substituting from equation 4.46 into equation 4.45 we get:
)1(22
2
DD
cc eff
−+=
γ
4.47
Dividing both numerator and denominator by γ, we get;
)1(22
1
1DD
cc eff
−+
=
γ
4.48
Both equations 4.47 and 4.48 represent the effective capacitance
between points A and B, Figure 4.6, where D is the duty cycle of the switch
in series with C1.
A plot of equation 4.47 is shown in Figure 4.7, for different values of γ.
Figure 4.7 Relation between D and Ceff for DCDS circuit as referred to branch C1.
If the duty cycle, D is referred to branch C2 and using equation 4.39; we can
get:
70
][1
2
2
2
^
^
)1( XDXDI
vX cc
AB
c eff
−+==
4.49
cD
cD
c eff 1
2
2
2 111)1(
ωωω⋅+⋅= −
4.50
cD
cD
c eff 1
2
2
2 111)1( ⋅+⋅= −
4.51
cc
cDcDc eff 21
2
2
1
2
)1(1 −+=
4.52
Taking the reciprocal of equation 4.52, we get:
cDcD
ccc eff
2
2
1
2
21
)1( −+=
4.53
Dividing by C2
)1(1 2
2
1
DD
cc eff
−+
=
γ
4.54
And dividing by γ
)1(22
2
DD
cc eff
−+=
γ
4.55
Equations 4.54 and 4.55 represent the value of the effective
capacitance in terms of the fixed capacitors C1 and C2, where D is the duty
cycle of the switch in series with C2.
A plot of equation 4.54 is shown in Figure 4.8, for different values of γ.
71
Figure 4.8 Relation between D and Ceff for DCDS circuit as referred to branch C2.
The feature of this topology is that the maximum effective capacitance
between points A and B, Figure 4.5, could be up to the sum of both fixed
capacitors (C1+C2), also it gives more flexibility to obtain the same value of
the effective capacitance at two different settings (a high and a low value) of
the duty ratio, due to the bell function shape of the relation, Figure 4.7and
Figure 4.8.
4.3 Implementation of the Switched-capacitance Circuit
The schematic diagram of Figure 4.9 shows a simple verification of a
switched capacitance circuit formed by two capacitors, each capacitor is
connected in series with two MOSFET switches and the two parallel
branches are connected to an inductor across a sinusoidal power supply.
72
Figure 4.9 Schematic diagram of the implemented switched capacitance circuit.
The two cases discussed above in this chapter have been implemented
in the Lab to verify the theoretical relations deduced above between the duty
cycle and the effective capacitance for the two cases, the single-capacitance,
double-switch case and the double-capacitance, double-switch case.
Figure 4.10 Experimental setup of the SCC.
The experimental system is setup as shown in Figure 4.10 which shows
the following components from right to left; the capacitor bank, the PC, the
DC power supply, the two switches with their drive circuits, the iron-cored
coil, multi-meters, the AC mains, zero crossing detector circuit, and the
microprocessor board respectively.
73
(a) (b)
Figure 4.11 (a) Drive circuit of the MOSFET switches. (b) Microprocessor board.
The switching signal is fed to the driver circuit of the power MOSFET
(STP11NK40Z) as shown in Figure 4.11(a), this switching signal is obtained
from the output port of a 8085 microprocessor board Figure 4.11(b), this
board is Kimberry Ltd product, model MT885. A simple assembly program is
written to generate a switching signal which has a duty cycle varying
between 0.1 and 0.9 while a zero detector circuit is used to detect the supply
frequency and synchronize it with the generated pulse from the
microprocessor port, the output of the zero detector circuit is supplied to the
input port of the microprocessor.
The power diode rectifier bridge shown in Figure 4.9 is used to rectify
the AC current through the MOSFET switch.
The duty cycle of the power switch is varied from 0.1 to 0.9 and the
voltage and current are measured, from which the capacitive impedance
could be calculated and hence the effective capacitance of the circuit
assuming that the ohmic resistance of the circuit is negligible.
The plot of the duty cycle against the effective capacitance for the
single-capacitor double-switch circuit is shown in Figure 4.12 for the value of
the single capacitor, C=10µF.
74
Figure 4.12 Experimental plot of D against Ceff for SCDS circuit.
Figure 4.13 Experimental plot of D against Ceff for DCDS circuit.
0
50
100
150
200
250
300
350
400
450
500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Duty Cycle, D
Eq
uiv
ale
nt
Ca
pa
cit
an
ce
(u
F) C2=C1
C2=2C1
C2=C1/2
0
200
400
600
800
1000
1200
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Duty Cycle, D
Eq
uiv
ale
nt
Ca
pa
cit
an
ce
Theoretical Curve
practical Curve
75
The plot of the experimental results which is obtained for the double-
capacitance double switch circuit are shown in Figure 4.13 for three different
cases which are comparable to the cases considered in Figure 4.7.
The deviation of the experimental characteristics of both circuits from
the theoretical characteristics is due to the fact that the capacitance is
calculated from the voltage and current readings which was not accurate due
to the gradual accumulation of the voltage across the capacitor which doesn’t
allow an exact reading of the this voltage, besides the approximation
assumed by neglecting the ohmic resistance of the circuit, but the
experiment however verifies the effect of the switching capacitance circuit
that acts as a variable capacitor between the terminals where the effective
capacitance is measured.
4.4 Other topologies of switched capacitance circuit
In the following sections, different topologies of the switched
capacitance circuit which were presented in [62] are shown in brief. It is to be
considered into account that the switched capacitance circuit could be used
in two different modes, the first mode is utilising the resonance energy
transfer concept, where the electrical stored energy in the capacitor is
interchanged with the magnetic stored energy in the motor inductor or
regenerated to the DC supply alternately, the second mode is utilising the
switched capacitance circuit as a variable capacitor.
76
4.4.1 Single-capacitance triple-switch circuit
Figure 4.14 Single-capacitor triple-switches.
The system shown in Figure 4.14 illustrates a single-capacitor, triple-
switch circuit which utilises the resonance energy transfer concept to
interchange energy between the capacitor, C and the motor inductor, L. The
capacitor C is charged via switch S2 by the energy stored in the motor
inductor, L. This capacitor, C, discharges its energy back to the motor
inductor via switches S1 and S3 at the beginning of the energizing period of
the motor phase, θ1 and θ5 in Figure 3.3.
Figure 4.15 Single-capacitor, triple-switches circuit with a coil.
In Figure 4.15, a single-capacitor, triple-switch system with an external
coil is illustrated. In this scheme, the resonance energy transfer concept is
77
applied as before. The energy stored in the inductance of the motor is
transferred to the capacitor C when the switch S3 is open (OFF) and the
switch S2 is closed (ON), this lasts for a time interval equals to 1/2π√(LC)
seconds at the beginning instant of the maximum inductance dead zone, θ2
in Figure 3.3. During the minimum inductance dead zone, θ4 in Figure 3.3,
switch S1 is closed for an interval equals to 1/2π√(L1C) seconds to transfer
the capacitor energy to the DC supply and the inductor L1. The inductor L1
acts as a current limiter against high current surge in switch S1 as well.
4.4.2 Single-capacitor 5-switches circuit
Figure 4.16 Single-capacitor 5-switches circuit.
Another system of a single-capacitor and 5-switches is shown in Figure
4.16 where the capacitor C is charged via switches S2 and S3 while
discharges its stored energy to the motor inductor via switches S1 and S4.
4.4.3 Hybrid switched capacitance systems
In the following, some switched capacitance circuit topologies are
shown, such systems are utilising both modes of operation, the resonance
energy transfer mode and the variable capacitance mode, hence the name
hybrid systems is given to such systems of switched capacitance circuits.
78
Figure 4.17 Single-capacitor triple-switches system with SCC.
Figure 4.18 Single-capacitor 5-switches system with SCC.
79
Figure 4.19 Single-capacitor 6-switches system with SCC.
In such systems, the variable capacitance effect of the SCC is used to
control the amplitude of the current while the resonance energy transfer is
used to improve the time response of the current waveform, in particular, the
rising and falling edges. Operation of various circuits is given in [62].
4.5 Applying the switching capacitance circuit to SRM drive topology
In the following chapter, the simulation results will be discussed for each
of the two above topologies, the SCDS and DCDS, and the validation of
these circuits will be proved and compared with the conventional asymmetric
converter from several points of view, including the effect of the switched
capacitance circuit on the rise and fall times of the current waveform, the
average torque, the harmonic content and the overall efficiency of the drive
system.
80
Chapter 5
Modelling and Simulation of the Proposed Drive Circuit
5.1 Introduction
Most studies concerning dynamic simulation of the switched reluctance
machine have been achieved by programming, also software designed to
simulate electric network systems as the EMTDC and EMTP have been
used. These techniques although very useful, have lack of flexibility [19].
However several simulation studies have been done using the so called
circuit-based languages such as SPICE, MATLAB Simulink or M-file,
MATRIX, VISSIM and MATHCAD to establish a model of the switched
reluctance machine. [11, 16-30, 63]
The SPICE models of the switched reluctance motor systems have the
advantage of accurate modelling of the electrical and the electronic
components of the drive circuit, but it is not capable of providing an accurate
model of the machine due to the fact that the magnetic circuits of the stator,
the air gap and the stators can’t be separated [64]. The MATLAB model of
the switched reluctance motor has the advantages of higher flexibility and
open environment which is capable of interacting with many other languages
such as C language as well as interface with digital electronic hardware such
as PIC Microcontrollers and DSPs which makes it more flexible in the control
modelling and processing, yet, the MATLAB environment with both the
Simulink and the M-file formats give an approximate model of both the
machine and the electronic and electrical components of the drive circuit.
However, few software packages have been developed recently to
simulate the switched reluctance machine accurately, such as the PC-SRD
developed by the speed laboratory in the university of Glasgow and the
Magnet software which is used in this research.
81
5.2 Magnet Software
The modelling and simulation work of this research has been carried out
using the Magnet software, which provides a friendly environment to model
the machine with its real mechanical and electrical parameters and computes
the electrical, magnetic, mechanical and thermal variables of the machine
taking into account the transient with motion effects, Figure 5.1 shows the
designed model of a 3-phase, 6/4 switched reluctance machine. For
simplicity and time saving, only half of the model is constructed as shown in
Figure 5.1 and Figure 5.3 as the other half of the machine will be a mirror
image of it.
Figure 5.1 Geometric Modeller of Magnet software.
The software supports 2D and 3D solid modelling construction of
components besides allowing the users to build their own library of user-
defined materials producing a wide and flexible variety of designs.
Magnet includes a circuit modeller, Figure 5.2 which permits the user to
connect the designed coils and power electronic devices to circuit drives and
loads in order to properly simulate the voltage and current present at the
82
terminals. It allows the selection of the coil material, the type whether solid or
stranded, the number of turns, the internal resistance besides any initial
conditions. The circuit components comply with SPICE standards.
Figure 5.2 Circuit modeller of Magnet software.
Once the geometrical model is produced and the drive circuit is
designed, the software utilises the finite element mesh analysis, as shown in
Figure 5.3 to calculate the magnetic characteristics of the machine as well as
the electrical and thermal characteristics. The limitation of this software with
respect to the research purpose, is that it takes very long time to run one
simulation for a full cycle, also it plots only one graph at a time with the real
time on the x-axis. For comparison reasons or to show the response with
respect to the position angle, the curve data has to be exported to excel and
re-plotted in terms of the desired parameter.
The Magnet has a plug-in that allows interaction with MATLAB which
makes it capable of achieving the advantages of both the SPICE model with
respect to the electronic components and the MATLAB model with respect to
flexibility and control schemes.
83
Figure 5.3 Finite element mesh of the solved model.
5.3 Parameters of the Simulated Machine
The parameters of the simulation model is taken from the 3-ph, 6/4
switched reluctance motor as in [48], this machine is a 0.5 kW with the
parameters shown in Table 5-1 below.
Table 5-1The machine parameters.
Stator outer diameter. Ds 134.3 mm.
Stator internal diameter. Dsi 114.3 mm.
Stator core axial length. Lcs 100 mm.
Stator core width. Cs 10 mm.
Rotor pole height. hr 13.35 mm.
Air gap. g 0.25 mm.
Shaft diameter. Dri 10 mm.
Stator pole arc. βs 40o
Rotor pole arc. βr 45o
Number of turns per phase. Nph 600
Winding wire diameter. Dw 0.5 mm.
84
The machine model is developed using the geometrical modeller as
described before with the true mechanical and electrical parameters shown
in the above table.
5.4 Drive Circuit
The drive circuit is based on the asymmetric converter circuit with the
switched capacitance circuit inserted in series with the phase windings as
shown in Figure 5.4.
Figure 5.4 Proposed drive circuit.
The two topologies of the switched capacitance circuits which are
discussed and analysed in chapter 4, are simulated here, the simulation
results of both circuits are compared with each other and with the results
obtained from the conventional asymmetric converter alone without inserting
the switched capacitance circuit.
For economical and size considerations of the converter circuit, the
other topologies of the switched capacitance circuits which include more than
two capacitances and/or two switches are not considered, as for each phase
there will be two capacitances and two power switches added to the original
asymmetric drive circuit besides the control circuit which may lead to a
significant increase in the size and volume of the converter, hence a
decrease in the power to volume ratio.
85
5.4.1 Asymmetric Converter Drive
The main aim of this chapter is to study the effect of the proposed
drive circuit on profiling the current wave shape and how it affects the output
torque as it is the control variable. It was stated before, in chapter 3, that the
principal of torque control of the switched reluctance motor depends on two
methods, either controlling the amplitude of the phase current or changing
the dwell angle. As a matter of fact, changing the dwell angle means in other
words allowing the phase current to flow for more time in order to develop
positive torque to reduce torque ripples. The proposed drive circuit has
shown an achievement of the advantages of both previous methods
together, as it allows the control of the amplitude of the phase current and
makes the current pulse more flat within the same dwell angle.
For this purpose, which is the study of the effect of the proposed
circuit on the current wave shape and the torque ripples, comparison is being
held between the output waveforms obtained from the conventional
asymmetric drive circuit, Figure 5.5, without inserting the switched
capacitance circuit and the waveforms obtained from operating the motor
under the same conditions after inserting the switched capacitance circuit.
86
Figure 5.5 Conventional Asymmetric Converter.
The conventional asymmetric drive circuit is setup as shown in Figure
5.5, with the motor parameters shown in Table 5-1.
5.4.1.1 Simulation results for the Asymmetric converter drive
For the study purpose, the motor is derived in two different modes of
operations, the single phase operation mode, which means only one phase
is energized at any instant while other phases are idle, and the two phase
operation mode, where phase two is energized at an instant when current is
still flowing in phase one. The single phase operation mode is used only to
examine the effect of the switched capacitance circuit on shaping the current
waveform. Simulation results of the single phase mode operation with dwell
angle of 24° is shown in Figure 5.6 to Figure 5.8 below.
ph
as
e 2
T1
T2
ph
as
e 3
T1
T2
ph
as
e 1
T1
T2
V1
PW
L
D1
S4
S5
D2
S3
D3
D4
S6
D5
S1
1
D6
S1
2
R1
17
R3
17
R4
17
87
Figure 5.6 Phase current profile for single phase operation mode.
Figure 5.6 shows the current pulse of the machine phase which is
energised at the instant t= 8 ms for a period of 4 ms, which corresponds to a
dwell angle of 24°. The current reaches its rated value of 2.3 A in 1.5 ms. By
definition, the rise time of this current pulse is 0.96 ms, which is the time
required for the current to increase from 0.23 A to 2.07 A. Similarly we can
find the fall time of this current pulse to be 3.5 ms. The phase voltage and
torque pulse of the same phase are shown in Figure 5.7 and Figure 5.8
respectively. These waveforms are to be compared with the corresponding
waveforms resulting from the operation of the SRM under the sae conditions
after inserting the switched capacitance circuit within the converter.
-0.5
0
0.5
1
1.5
2
2.5
3
6 8 10 12 14 16
Time (ms)
Cu
rre
nt
(A)
88
Figure 5.7 Voltage across the motor phase.
Figure 5.8 Torque pulse developed by a single phase.
The torque pulse of Figure 5.8 has a peak of 3.54 Nm.
In the following section, the new drive circuit topology is presented
using the single-capacitance double-switches, switched capacitance circuit.
-300
-200
-100
0
100
200
300
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Vo
ltag
e (
V)
-1
0
1
2
3
4
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
To
rqu
e (
N.m
)
89
5.4.2 Single Capacitance Double Switch (SCDS) drive Circuit
The switched capacitance circuit is inserted in series with the motor
phase, the circuit consists of a single capacitance connected in series with a
semiconductor switch to form one branch which is connected in parallel with
another semiconductor switch as shown in Figure 5.9 below. The 17 Ω
resistor represents the phase ohmic resistance measured in section 3.5.1.1.
The value of the fixed capacitors, C1, C2 and C3 is selected to be 60µF
as will be explained below.
Figure 5.9 Drive circuit utilising single-capacitance double-switch circuit.
5.4.2.1 Switching Strategy
In order to hold a valid comparison between the current case where the
single-capacitance double-switch switched capacitance circuit (SCC) is
inserted in series with the motor phase, and the previous case where the
conventional asymmetric converter is used, the switching pattern of the main
switches of each phase (S3, S4, S5, S6, S11 and S12) is kept the same, while
the switching pattern of the switched capacitance circuit switches (S1, S2, S7,
S8, S9 and S10) is designed so that the resulting effective capacitance of the
90
circuit tracks the variation of the phase inductance profile to resonate with it
at the supply frequency.
5.4.2.2 Calculation of the effective capacitance, Ceff
From the above discussion and for a given value of the phase
inductance, L at any instant, the relation between L and Ceff can be written as
stated in equation 5.1 below which represents the resonance frequency of an
RLC series circuit:
CLf
effπ2
1
0=
5.1
Where; f0 is the supply frequency in our case, from which, Ceff could be
found from equation 5.2 as:
fLC eff 2
0
1
2 π=
5.2
5.4.2.3 Calculation of Duty Cycle, D
For the single capacitance double switch SCC, the duty cycle is
calculated straight forward from equation 4.33 to be as stated in equation 5.3
below.
C
CD
eff
=
5.3
In the above equation, in order to calculate for D, the value of C, which
is the fixed capacitor, is the only unknown as Ceff is calculated as above. So,
the switching strategy is based on the estimation of the value of L at each
instant of the operation, from which the corresponding value of Ceff is
calculated, hence the duty cycle D which is continuously varying with the
variation of the phase inductance.
91
5.4.2.4 Determining the value of the fixed capacitor, C
The value of the fixed capacitor, C is a factor of determining the duty
cycle at which the semiconductor switch is operated, although it is arbitrary,
there has to be a minimum value of C so that the duty cycle is not less than
10% of the period of the switching function, so from equation 5.3 the value of
the fixed capacitor, C can be found as follows:
)(max
min CC
Deff
=
5.4
DCC eff
2
minmax)( ⋅= 5.5
According to the maximum value of the effective capacitance which is
366 µF (this is the value of Ceff which resonates with the minimum
inductance) and setting Dmin to 0.1 the exact value of C could be found, this
value will be the maximum effective capacitance which could be obtained
between the terminals XX of Figure 4.1 at D=0.1. The max value of Ceff
according to Figure 4.4 is the value of the fixed capacitor, C divided by D2
which is 0.01 in this case. As the value of C is arbitrary, so it is selected to be
60 µF as this value will require a duty cycle greater than 0.15 to get the 366
µF and the maximum value of D will be around 0.9 which will be more
convenient for practical implementation.
5.4.2.5 Simulation results for SCDS drive circuit
The corresponding waveforms for those shown above are presented
below for the phase current, voltage and torque pulse at the transient
(starting) period.
In this case, the switching frequency of the switched capacitance circuit
is varied over a wide range in order to study its effect on the current profile
and on the overall performance of the drive circuit, the shown waveforms are
for a switching frequency of 1kHz as it is recommended in [61] to keep the
ratio of the switching frequency to the supply frequency greater than 10 in
order to reduce the switching harmonics. Study has been carried out for a
wide range of switching frequencies between 300 Hz up to 100 kHz with 0.5
92
kHz increment in order to examine the effect of the switching frequency on
various characteristics of the motor and drive circuit.
Figure 5.10 Phase current profile for SCDS circuit, f=1kHz.
Figure 5.11 Voltage across the motor phase for SCDS circuit, f=1kHz.
-1
0
1
2
3
4
5
6
6 8 10 12 14 16
Time (ms)
Cu
rre
nt
(A)
-300
-200
-100
0
100
200
300
400
500
6 8 10 12 14 16
Time (ms)
Vo
lta
ge
(V
)
93
Figure 5.12 Torque pulse developed by a single phase for SCDS circuit, f=1kHz.
As it can be seen from the above figures of the phase current profile,
Figure 5.10, the phase voltage, Figure 5.11, and the torque pulse developed
per phase, Figure 5.12, some significant effects of the switching capacitance
circuit can be found compared to the case of the asymmetric converter drive
circuit on both the current shaping and the torque pulse consequently, the
effect of the switched capacitance circuit on the phase current and the
developed torque is discussed below.
(a) Effect of SCC on the phase current profile
Figure 5.13 shows the combination of the two current waveforms
resulting from the simulation of the two cases, these are: the asymmetric
converter without the SCC, and the same converter with the SCC inserted in
series with the motor phase, setting the switching frequency of the SCC to
1kHz. From Figure 5.13 the following points can be obviously found out:
-1
1
3
5
7
9
11
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
To
rqu
e (
N.m
)
94
Figure 5.13 Current profiles for the asymmetric converter with and without SCC.
1. The amplitude of the current pulse is increased in the second
case to 4.79 A instead of 2.3 A in the first case, which represents
108% increase of the amplitude of the starting current.
2. The current pulse starts rising up at the instant t=7ms instead of
8 ms (the exact switching instant of phase 1), which in other
words is equivalent to an advance angle of 6°.
3. Comparing the two current profiles show that the new technique
allows the control of the phase current profile to give the effect of
controlling of both the dwell angle and the amplitude of the phase
current. The dwell angle is a factor of optimizing the efficiency,
while the current amplitude is a factor of maximizing the
developed torque. Actually the dwell angle in the second case is
the same as the first case, but the flat top pulse with the fast rise
and fall times gives the same effect as if the dwell angle has
been increased.
-1
0
1
2
3
4
5
6 8 10 12 14 16
Time (ms)
Cu
rre
nt
(A)
with SCC,f=1kHz
without SCC
95
(b) Effect of SCC on the developed torque
Considering Figure 5.14 below, which represents the two torque pulses
resulting from the two above cases, the significant effect on the torque pulse
is clearly shown under the same operating conditions which is due to the fact
that the torque developed by the switched reluctance machine is actually
proportional to the square of the phase current (i2). This fact leads to make
the effect of any slight change of the current profile more significant on the
developed torque as could be seen in Figure 5.14. The torque pulse in the
case of using the SCC starts to build up exactly at instant t=8 ms or slightly
before it, instead of starting at t=8.5 ms in the first case. Also, the flat top of
the torque pulse means an increase of the average value of the developed
torque and hence the overall efficiency of the drive system. Finally it could be
noted that the peak value of the torque pulse is increased from 3.54 Nm to
10.58 Nm which is equivalent to 198.9 % increase.
Figure 5.14 Developed torque for the asymmetric converter with and without SCC.
-0.5
1.5
3.5
5.5
7.5
9.5
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
To
rqu
e (
Nm
)
with SCC,f=1kHz
without SCC
96
(c) Effect of SCC on the voltage stresses
The effect of using the switched capacitance circuit on the phase
voltage could be seen in Figure 5.15 below, which shows the voltage being
chopped during the ON period at the switching frequency (1kHz) in this case.
The voltage chopping increases the voltage stress over the motor phase to
271 V (RMS) which represents 23% more than the asymmetric converter
case before inserting the switched capacitance circuit. However, this
increase of voltage across the machine phase doesn’t reflect on the voltage
stress across the upper semiconductor switch as it occurs only between the
instants 8 ms and 12 ms, while the switch is ON, this can be clearly seen in
Figure 5.16.
Figure 5.15 Voltage stress over the machine phase.
-400
-300
-200
-100
0
100
200
300
400
500
6 7 8 9 10 11 12 13 14 15
Time (ms)
Vo
lta
ge (
V)
with SCC,f=1kHz
without SCC
RMS value
97
Figure 5.16 Voltage stress across the upper switch, S4.
(d) Effect of SCC on the current harmonics
The effect of using the switched capacitance circuit is studied for the
previous case which is the single-capacitor double-switch circuit with a
switching frequency of 1kHz compared with the case of the conventional
asymmetric converter without using the switched capacitance circuit.
-230
-200
-170
-140
-110
-80
-50
-20
10
6 7 8 9 10 11 12 13 14 15
Time (ms)
Vo
ltag
e (
V)
98
Figure 5.17 Respective Harmonic Amplitudes.
The respective amplitudes of odd harmonics in both cases are shown in
Figure 5.17 while the total harmonic distortion factor (THD) is shown in
Figure 5.18 for both cases. However, the THD decreases with increasing the
switching frequency according to the profile shown in Figure 5.19.
Figure 5.18 Total Harmonic Distortion (THD).
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
Ha
rmo
nic
Am
pli
tud
e
i1 i3 i5 i7 i9 i11 i13 i15 i17 i19 i21
Harmonic Order
without SCC
with SCC, f=1kHz
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
To
tal
Harm
on
ic d
isto
rtio
n (
TH
D)
without SCC with SCC, f=1kHz
99
Figure 5.19 Variation of THD with switching frequency.
However, the calculation of the THD is not very accurate as it has been
calculated using a semi-analytical semi-numerical method [63] which needs
at least 200 samples to be read which was not available in some cases due
to technical issues with the software, also some of the undesired harmonic
orders could be eliminated, but it is not in the scope of this research.
(e) Effect of the SCC on the overall efficiency
Referring to equations 3.8 and 3.9, the mechanical output of the motor could be written as:
dt
wdRivi
dt
wd f
m
m−−=
2 5.6
That is, taking the ohmic losses and the magnetic stored power away of
the electrical input power. Hence, by determining the ohmic losses and the
magnetic energy in both cases (the one of the asymmetrical converter and
the one with the switched capacitance circuit), comparison of the effect of
using the switched capacitance circuit on the overall efficiency of the system
can be achieved.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
1 11 21 31 41 51 61 71 81
Switching frequeny (kHz)
TH
D f
acto
r
100
Figure 5.20 Ohmic Losses per coil.
The ohmic losses profiles in both cases as obtained from the simulation
results are shown in Figure 5.20, where the shown ohmic loss is per coil of
the stator winding (this value has to be multiplied by 2 to find the ohmic
losses per phase) while the instantaneous magnetic energy profiles are
shown in Figure 5.21.
Substituting the average values of the ohmic losses ( i2Rm), and the
magnetic power (dwf/dt) into equation 5.6, the overall efficiency, η is found to
be η=93.06% for the asymmetric converter circuit and η=91.43% for the
switched capacitance converter circuit.
-1
1
3
5
7
9
11
6 8 10 12 14 16
Time (ms)
Oh
mic
Lo
sses (
W)
with SCC, f=1kHz
without SCC
101
Figure 5.21 Instantaneous Magnetic Energy.
(f) Effect of the SCC on the switching losses
The reduction in the overall efficiency can be referred to the increased
harmonic content of the phase current and the switching losses in the
switched capacitance circuit. The contribution of the switched capacitance
circuit to the switching losses is carried out by comparing the switching
losses in the two cases, the asymmetric converter without the SCC and the
case with the use of the SCC at 1kHz, bearing in mind that in the first case,
nearly all of the power dissipated in the MOSFETs occurs when the device is
in the ON state [65] and this is included in the ohmic losses.
The switching losses, Ps can be calculated to a good approximation
using equation 5.7 below. [66]
τfIVP s
mm
s 2= 5.7
Where; Vm is the maximum voltage stress on the switch; Im is the
maximum current through the switch; fs is the switching frequency; and τ is
-0.1
0.1
0.3
0.5
0.7
0.9
1.1
1.3
1.5
6 8 10 12 14 16
Time (ms)
Mag
neti
c E
nerg
y (
J)
with SCC,f=1kHz
without SCC
102
the period of the switching interval, which is the sum of the delay time plus
the rise time specified in the device datasheet.[64]
Equation 5.7 is applied to the current case study with referring to Figure
5.22 below, which shows the voltage stresses over the two power switches
associated with the switched capacitance circuit at a switching frequency of
1kHz, from which τ can be estimated, hence the switching losses can be
calculated. A compromise has to be done between the harmonic content
which decreases with increasing the switching frequency and the switching
losses which increases with increasing the switching frequency.
Figure 5.22 Voltage stresses over S1 and S2 at 1kHz.
5.4.2.6 Effect of the Switching Frequency
The effect of the switching frequency of the switched capacitance circuit
has been studied over a wide range of switching frequencies between 300
Hz and 100 kHz in order to examine the effect of the switching frequency on
profiling the phase current and its influence on the developed torque, the
harmonic distortion, the switching losses hence the overall efficiency of the
drive system.
-250
-200
-150
-100
-50
0
50
100
150
200
250
8 9 10 11 12
Time (ms)
Vo
ltag
e (
V)
ττττ
S1
S2
103
The following figures from Figure 5.23 to Figure 5.34 show the different
profiles of phase current for switching frequencies between 300 Hz and 100
kHz at the same dwell angle of 24° as stated earlier. The current profiles are
compared with the asymmetric converter current profile, which is shown in
dotted line in all graphs.
Figure 5.23 Current profiles for 300 Hz and 500 Hz.
At 300 kHz the current waveform is as shown in Figure 5.23, the
simulation results showed high current and voltage spikes over the motor
phase after the second switching interval at this frequency, the voltage spike
reached few kilovolts which led to a significant deformation of the magnetic
characteristics (the flux linkage, the magnetic energy, and the magnetic co-
energy) of the machine, however, it was recommended in [61] not to use a
switching frequency less than 10 times the supply frequency which means a
switching frequency of 500 Hz minimum.
-0.5
0.5
1.5
2.5
3.5
4.5
6 8 10 12 14 16
Time (ms)
Ph
as
e C
urr
en
t (A
)
f=500 Hz
f=300 Hz
104
The full details of the case of 1 kHz have been shown above as it was
the case of lowest switching losses, with a relatively high peak and good time
response.
At 1.5 kHz, the simulation results showed lower peak of the current
waveform than that at 1 kHz, with higher switching losses and a negative flux
linkage-peak of about 0.35 Wb as shown in Figure 5.24. Also, the simulation
results showed high voltage spikes (10 kV) over the switches S1 and S2
under the same operating conditions as the 1 kHz case.
Figure 5.24 Flux linkage at 1.5 kHz.
The results of switching frequencies between 2 kHz and 7 kHz have
shown good current and voltage profiles of the motor phase with normal
operating conditions of the power electronic switches, the limitation of this
range of frequencies is the high switching losses shown in Figure 5.35.
There has been an unexpected failure of the system at a frequency of
7.5 kHz.
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
6 8 10 12 14 16 18 20 22
Time (ms)
Flu
x L
inkag
e (
Wb
)
f=1.5 kHz
without SCC
105
Figure 5.25 Current profiles for 1kHz and 1.5 kHz.
Figure 5.26 Current profiles for 2 kHz, 2.5 kHz and 3 kHz.
-0.5
0.5
1.5
2.5
3.5
4.5
6 8 10 12 14 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=1 kHz
f=1.5 kHz
-0.5
0.5
1.5
2.5
3.5
4.5
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=2.5 kHz
f=2 kHz
f=3.5kHz
106
Figure 5.27 Current profiles for 4 kHz and 5 kHz.
Figure 5.28 Current profiles for 5.5 kHz and 6 kHz.
-0.5
0.5
1.5
2.5
3.5
4.5
5.5
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=4 kHz
f=5 kHz
-0.5
0.5
1.5
2.5
3.5
4.5
5.5
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=5.5 kHz
f=6 kHz
107
Figure 5.29 Current profiles for 6.5 kHz and 7 kHz.
The examination of the behaviour of the converter-motor system in the
switching frequency range between 8 kHz and 30 kHz has shown the current
profiles as in Figure 5.30 and Figure 5.31 below, with normal voltage
stresses and peak currents, while it has shown maximum increase in the
output mechanical energy at a switching frequency of 20 kHz as shown in
Figure 5.47.
The switching frequency range above 40 kHz has proved to have very
slow time response of the current waveform in most cases as could be seen
from figures below, besides the extremely higher switching losses in the
power electronic switches and a significant deformation of the magnetic
characteristics of the machine which led to a very low value of the average
developed torque due to negative torque pulses and a very poor overall
efficiency due to the extremely high switching losses.
-0.5
0.5
1.5
2.5
3.5
4.5
5.5
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=7 kHz f=6.5 kHz
108
Figure 5.30 Current profiles for 8.5 kHz and 10 kHz.
Figure 5.31 Current profiles for 20 kHz and 30 kHz.
-0.5
0.5
1.5
2.5
3.5
4.5
5.5
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=10 kHz f=8.5 kHz
0
2
4
6
8
10
12
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=20 kHz
f=30 kHz
109
Figure 5.32 Current profiles for 40 kHz and 50 kHz.
Figure 5.33 Current profiles for 60 kHz and 70 kHz.
0
2
4
6
8
10
12
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=40 kHz
f=50 kHz
0
2
4
6
8
10
12
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=70 kHz
f=60 kHz
110
Figure 5.34 Current profiles for 80 kHz and 100 kHz.
The effect of changing the switching frequency of the switched
capacitance circuit is obviously readable from the above figures, which show
the change of the current profile with changing the frequency including the
amplitude, the rise and fall times, the flatness of the top, and the advance
angle, however the comments and detailed calculations are shown for a case
study of the 1 kHz frequency as stated in section 5.4.2.5 (a).
The relation between the different parameters of the current profile and
the switching frequency of the switched capacitance circuit could not be
represented by neither a mathematical nor an empirical relation up to the end
of this research, but it will be an interesting field of research for the future
work.
The significant parameter according to which the switching frequency
could be decided is the switching losses as if it goes to high levels with
increasing the switching frequency it will severely affect the overall efficiency
of the system besides the increased size and cost of the heat sinks
associated to overcome the problems of overheating of the power
semiconductor switches.
0
0.5
1
1.5
2
2.5
3
3.5
6 7 8 9 10 11 12 13 14 15 16
Time (ms)
Ph
ase C
urr
en
t (A
)
f=100 kHz
f=80 kHz
111
Figure 5.35 below shows the calculated switching losses for each
switching frequency within the studied range.
Figure 5.35 Switching losses in terms of the switching frequency.
According to the calculations of the switching losses and as it could be
seen in Figure 5.35, the minimum switching losses occurs at the frequency
fs=1 kHz, hence it is selected as a case study for the comparisons and
discussions above. The frequencies above 10 kHz are neglected as they
have shown unacceptable levels of the switching losses compared with the
power level of the drive system under consideration.
5.4.3 Two Phase Operation
The above discussion has been considering the single phase operation
of the SRM in order to compare the influence of inserting the switched
capacitance circuit on different parameters of the drive system, the single
phase operation means that one phase of the machine is energised at a
time, while the other phases are idle. But from the practical point of view, the
single phase operation mode leads to high torque ripples which contribute to
the acoustic noise, mechanical vibration, and velocity errors due to the
0
20
40
60
80
100
120
140
160
180
200
0 1 2 3 4 5 6 7 8 9 10 11
Switching Frequency (kHz)
Sw
itc
hin
g L
os
ses/p
ha
se (
W)
112
torque pulsation during the operation of the machine, therefore the two or
multi-phase operation is recommended as one factor to reduce the torque
ripples, the other factors are minimizing the hysteresis band of the current
control loop and keeping the dwell angle constant over the operation range.
Therefore, simulation is repeated under the same conditions as before
while allowing the phase currents to overlap.
The maximum limit of the overlap angle, θomax, could be calculated from
equation 5.8 below. [1]
)1
21
(2
max
qP r
o−−=
πθ
5.8
Where; Pr is the number of rotor poles and q is the number of phases.
The maximum overlap angle according to equation 5.8 is found to be 15°.
However, the overlap angle is adjusted to 9° and the simulation results
for this case are shown in the following figures, where Figure 5.36 and Figure
5.37 are showing the phase currents and the developed torque respectively
for the asymmetric converter without the SCC circuit while Figure 5.38 and
Figure 5.39 are showing the phase currents and the developed torque
respectively for the converter with the SCC at a switching frequency of 1
kHz. The significant factor to be monitored in this case is the influence of
using the switched capacitance circuit on the torque ripples.
113
Figure 5.36 Phase Currents in case of overlap, without SCC.
Figure 5.37 Developed torque in case of overlap, without SCC.
Without SCC
-0.1
0.4
0.9
1.4
1.9
2.4
2.9
0 5 10 15 20
Time (ms)
Ph
as
e C
urr
en
ts (
A)
9°°°°
Phase 1 Phase 2 Phase 3
-30
-20
-10
0
10
20
30
0 10 20 30 40 50 60
Time (ms)
Avera
ge T
orq
ue (
Nm
)
114
Figure 5.38 Phase Currents in case of overlap, using SCC.
Figure 5.39 Developed torque in case of overlap, using SCC.
With SCC, f=1kHz
-0.1
0.9
1.9
2.9
3.9
4.9
5.9
0 5 10 15 20
Time (ms)
Ph
as
e C
urr
en
ts (
A)
Phase 1 Phase 2 Phase 3
0
0.2
0.4
0.6
0.8
1
0 10 20 30 40 50 60
Time (ms)
Avera
ge T
orq
ue
(N
m)
115
Referring to Figure 5.37 and Figure 5.39, where the average developed
torques by the three-phase motor are shown in the continuous torque
development mode, it could be seen obviously by comparing the two figures
the effect of using the switched capacitance circuit in reducing the torque
ripples at steady state. To calculate the torque ripples in the first case of
Figure 5.37, the steady state period is zoomed in Figure 5.40, which shows a
torque ripple (TR) of 2.15 Nm measured from the average value of the
developed torque of 0.65 Nm.
Figure 5.40 Torque ripples in asymmetric converter.
The following figures from Figure 5.41 to Figure 5.45 show the average
torque developed by the switched reluctance motor for different switching
frequencies of the switched capacitance circuit, the average torque is shown
in dotted line in all figures.
0
0.5
1
1.5
2
2.5
3
3.5
30 35 40 45 50 55 60
Time (ms)
Avera
ge T
orq
ue (
Nm
)
Average
value
TR
116
Figure 5.41 Developed torque at f=30 kHz.
Figure 5.42 Developed torque at f=40 kHz.
-1
-0.5
0
0.5
1
1.5
2
0 10 20 30 40 50 60
Time (ms)
Avera
ge T
orq
ue
(N
m)
-2
-1
0
1
2
3
4
5
0 10 20 30 40 50 60
Time (ms)
Avera
ge T
orq
ue (
Nm
)
117
Figure 5.43 Developed torque at f=70 kHz.
Figure 5.44 Developed torque at f=80 kHz.
-1
-0.5
0
0.5
1
1.5
2
0 10 20 30 40 50 60
Time (ms)
Avera
ge T
orq
ue
(N
m)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 10 20 30 40 50 60
Time (ms)
Av
era
ge
To
rqu
e (
Nm
)
118
Figure 5.45 Developed torque at f=100 kHz.
The simulation results have shown a maximum torque ripple of 0.13 Nm
which occurs at a switching frequency of 100 kHz compared with a peak
torque ripple of 2.15 Nm in the case of the asymmetric converter before
inserting the switched capacitance circuit.
The change in the magnetic characteristics of the switched reluctance
motor due to the use of the switched capacitance driver circuit is shown in
Figure 5.46 which indicates the increase in the output mechanical energy
represented by the area enclosed by OABCO in the asymmetric converter
case and the area enclosed by OA1B1C1O in the switched capacitance
converter case.
-0.05
0.05
0.15
0.25
0.35
0.45
0.55
0.65
0 10 20 30 40 50
Time (ms)
Avera
ge T
orq
ue (
N.m
)
119
Figure 5.46 Flux Linkage versus Stator Current (Simulated).
The maximum increase in the output mechanical energy occurred at a
frequency of 20 kHz under same operating conditions as shown in Figure
5.47 below.
Figure 5.47 maximum increase in output energy at 20 kHz.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.5 1 1.5 2 2.5 3 3.5
Current (A)
Flu
x L
inkag
e (
Wb
)
without SCC
with SCC,f=1kHz
A
BC
A1
B1
C1
O
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
0 1 2 3 4 5 6
Current (A)
Flu
x L
ink
ag
e (
Wb
)
f=20kHzwithout
SCC
120
5.4.4 Double-Capacitance Double-Switch Circuit
The second topology of the switched capacitance circuit considered in
this research is the double-capacitance double-switch circuit shown in Figure
5.48 below.
Figure 5.48 Double-Capacitance Double-Switch Circuit.
The double-capacitance double-switch circuit utilises two fixed
capacitors instead of one in the case of single-capacitance double-switch
circuit. The same procedure as the previous case is followed to calculate the
effective capacitance and the values of the two fixed capacitors.
5.4.4.1 Determining the values of C1 and C2
Referring to Figure 4.7, it is shown that the relation between the
effective capacitance, Ceff and the duty cycle, D in the case of double-
capacitance double-switch circuit is a second order relation, which means
there will be two possible values of D, for each value of Ceff.
As has been done in the case of single-capacitance double-switch
circuit, the value of the fixed capacitor was calculated based on the value
which resonates with the minimum motor inductance at the supply frequency.
For the double-capacitance double switch case, the value of the capacitor C1
phase 2
T1
T2
phase 3
T1
T2
phase 1
T1
T2
V1 PWL
C1 C2
S1 S2
D1
S4
S5
D2
S3
D3
S7 S8
D4
C3 C4
S6
D5
S9 S10
C5 C6
S11
D6
S12
R1 17 R3 17 R4 17
121
is calculated to be less than or equal to the value which resonates with the
maximum value of the inductance of the motor phase (this is the minimum
value of the effective capacitance), while the value of the fixed capacitor C2
is calculated so that the sum of C1+C2 is greater than or equal to the value of
the capacitance which resonates with the minimum value of the phase
inductance (this value is the maximum value of the effective capacitance).
The calculated values of the capacitances which resonate with the
minimum and maximum values of the phase inductance at the supply
frequency are 100 µF and 6 µF respectively.
5.4.4.2 Calculation of the duty cycle
Referring to equation 4.47 which states the relation between the
effective capacitance, Ceff and the duty cycle, D, this equation can be re-
written as follows:
02
22
)1( =−+ − CDCDC effeff γ 5.9
02
2 )21(2
=+ −+− CDDCDC effeff γ
5.10
02
22
2=+ −+− CDCDCCDC effeffeffeff γ
5.11
02 )(2)1(
2=+ −+−+ CCDCCDC effeffeffeff γ
5.12
Equation 5.12 is a second order equation which can be solved for D,
provided that the effective capacitance, Ceff and the values of the fixed
capacitors C1 and C2 are known.
A simple program is written to solve equation 5.12 for each
instantaneous value of Ceff to get the corresponding value of the duty cycle,
D and provide it to the simulation software.
122
5.4.4.3 Simulation results of the DCDS drive circuit
Figure 5.49 DCDS circuit for f=1 kHz.
Figure 5.49 shows the current profile resulting from the double-
capacitance double-switch (DCDS) circuit compared with the current profile
of the asymmetric converter without using the SCC.
The current in the case of the DCDS circuit at a switching frequency of
1 kHz reaches a peak of 14.5 A with a very bad characteristics with respect
to the rise time and fall time.
The DCDS circuit didn’t show encouraging results along the range of
frequency between 1 kHz and 10 kHz, which is shown in Figure 5.50. As a
matter of fact, the low controllability of the DCDS circuit is referred to the fact
that there is always one capacitor at least has to be connected in series with
the motor phase, unlike the SCDS case, where the capacitor can be expelled
out of the loop by switching off the electronic switch in series with it, reverting
the drive circuit back to the case of the asymmetric converter circuit, this
gives more controllability to the SCDS circuit which is not available in the
DCDS case.
-0.5
1.5
3.5
5.5
7.5
9.5
11.5
13.5
15.5
6 8 10 12 14 16 18 20 22 24 26
Time (ms)
Cu
rren
t (A
)
DCDS
circuit,
f=1kHz
without SCC
123
The results have shown high current peaks with poor time-
characteristics, high negative torque peaks, and deformation of the voltage
and flux linkage of the motor phases besides an ohmic loss which is 5 times
more than the SCDS drive circuit.
Not enough study of the double-capacitance double-switch circuit has
been conducted in this research due to the expiration of the licence of the
software used in carrying out the simulation, but it might be of interest in the
future work.
Figure 5.50 DCDS circuit for f=10 kHz.
-1
0
1
2
3
4
5
6
7
8
6 8 10 12 14 16 18 20 22 24 26
DCDS
circuit,
f=10kHz
without SCC
124
Chapter 6
Conclusions and Future Work
6.1 Conclusions
This study has introduced the switched capacitance circuit as a new
topology of a drive circuit of a three phase, 6/4 switched reluctance machine
which hasn’t been covered by any published research until the time when
this thesis was completed.
The literature work considering the switched reluctance machine drive
circuits has been critically reviewed in both the design of the switched
reluctance machine and the design of its drive circuit topologies. Comparison
of the most common drive circuit topologies has been held showing the
major advantages and disadvantages of each topology together with its
major field of application. The literature review showed that no research has
been conducted on the application of the switched capacitance circuit in the
field of switched reluctance machine drive circuits.
The analysis and mathematical evaluation of the switched capacitance
circuit as a variable capacitor has been developed for two cases, which are
the single-capacitance double-switch case and the double-capacitance
double-switch case which have been introduced as drive circuits of the
switched reluctance motor in this research and the relation between the
effective capacitance and the duty cycle has been derived for both cases.
Other topologies of the switched capacitance circuits which utilise more
than two switches have been introduced in brief, besides those topologies of
the switched capacitance circuits based on resonance energy transfer
concept.
125
The latter two topologies were not discussed in details for economical
and design considerations of the drive circuits, as for example, if the single-
capacitor, triple-switches switched capacitance circuit is considered as a
drive circuit of a three-phase switched reluctance motor, this will add extra
nine switches and three capacitors to the conventional drive circuit with their
drive circuits and heat sinks which will contribute significantly to the increase
in size and weight of the drive system, besides the increase in the switching
losses associated. For these reasons, only the single-capacitance double-
switch and the double-capacitance double-switch were considered in details.
The mathematical derivation of the characteristics of both topologies
proved the behaviour of the switched capacitance circuit as a variable
capacitor in direct relation with the duty cycle of the power electronic switch
which gives an advantage of the new drive circuit topology which uses the
switched capacitance circuit as a variable capacitor to track the variation of
the phase inductance enabling a significant change of the phase current
profile with respect to the fall and rise times, the pulse duration, and the
amplitude.
In the literature, each one of the previous parameters of the phase
current pulse is controlled separately through certain techniques to meet the
requirement of a certain application, for example, to maximise the developed
torque, it is required to increase the dwell angle, which in contradiction,
decreases the overall efficiency of the drive system, also for optimum
operation, the dwell angle may be changed which leads to higher torque
ripples.
On the other hand, hysteresis current controller is usually used to keep
the current pulse flat topped.
The new technique of the switched capacitance circuit enables the
control f the phase current pulse to change its amplitude, its flat top, and its
rise and fall times all in one technique which gives an important advantage to
this research.
126
The characteristics of the switched capacitance circuit as a variable
capacitor has been verified experimentally and the characteristic curves have
been compared with the theoretical ones for both cases of the switched
capacitance circuits discussed in the theoretical part of the study. The
comparison showed an excellent agreement between the theoretical and the
practical behaviour of the switched capacitance circuit as a variable capacitor
if the reading errors and the measuring instruments tolerance are taken into
consideration.
The switched capacitance circuit has been imposed to the conventional
asymmetric converter as it is the most common converter topology which is
used in many SRM applications. It has been found to be an effective way of
shaping the phase current profile which gives a significant advantage of this
research. The literature showed that the principal of torque control of the
switched reluctance motor depends on one of two methods, either controlling
the amplitude of the phase current or changing the dwell angle. The
simulation results has proven that the switched capacitance circuit allows
control over both the current amplitude and the dwell angle both in one
technique.
By reviewing the different software programmes used to model and
simulate the SRM drive systems, there were few alternatives to be used in
this research, these were Pspice, MATLAB or one of the special purpose
software packages which have been recently developed. At the early stage
of the research, a full model of the motor with its conventional asymmetric
converter has been constructed using MATLAB –Simulink, few figures of this
model is shown in the appendix.
The limitation of the MATLAB model is represented in the following
major points:
• The switched reluctance machine model on MATLAB is idealised
specially with respect to the phase inductance where it is
represented by five linear segments summed together to give the
ideal profile of the phase inductance, this might be a good
approximation for a controller design, but in our case, the
inductance value was very important to be accurate as it is the
127
main parameter used to decide the value of the effective
capacitance to resonate with it at the supply frequency.
• The power electronics part of the drive system couldn’t be
modelled on MATLAB, such as the power supply, and the
converter circuit. Also, there was a necessary need to access the
component levels of the drive circuit to examine the effect of
different operating conditions on each passive and power
electronic component in the circuit.
The second alternative was the Pspice, but due to the fact that it is only
few researches in the literature which used Pspice to model a switched
reluctance machine no serious attempts were undertaken to use it in this
research.
The third alternative was the special purpose software packages
specially developed to model and simulate the switched reluctance motor
with its full drive and control system.
As a first option the PC-SRD software developed by professor T. J.
Miller and his team in Glasgow University was selected, but due to budget
limitations, it was not possible to use it in this research.
The SRM with its complete drive circuit has been modelled utilising the
Magnet software which allowed accurate modelling of the machine geometry
and magnetic characteristics besides the PSPICE-complying models of the
passive components and power electronic switches of the drive circuit
provided by its circuit modeller.
The simulation results of the single-capacitance double-switch drive
circuit has shown a very encouraging results which allows control over the
phase current wave shape and peak, which contributes significantly to the
mechanical output energy of the motor and hence the overall efficiency as
well as a significant enhancement in the torque ripples.
Simulation of the single-capacitance double-switch circuit has been
carried out to for a wide range of switching frequencies in an attempt to
achieve an empirical relation between the switching frequency and the
different significant parameters of the phase current profile.
128
However, to decide the optimum switching frequency of the switched
capacitance circuit, a compromise has to be done between the switching
losses which increase with increasing the switching frequency and the total
current harmonic distortion which decreases with increasing the switching
frequency. The use of a current limiter might be a good solution of this
problem.
The single-capacitance double-switch converter circuit would be a
perfect solution for SRM applications which require a high starting torque,
where the capacitor might be inserted in the circuit at the starting transient
period, then it could be expelled out of the circuit during normal operation.
Also, it is suitable for high speed applications due to its positive effect
on building up the phase current and improving its transient response at both
the turn-on and turn-off periods.
The double-capacitance double-switch drive circuit didn’t show
encouraging results compared with the single-capacitance one, as it has
shown high current peaks with negative torque peaks which significantly
reduces the average developed torque besides the low controllability due to
the permanent existence of at least one capacitor connected in series with
the machine phase. However, more research is required to be conducted on
such a topology for more validation.
6.2 Future work
The following points might be of great interest for future work related to
the utilisation of switched capacitance circuit as a drive system of the
switched reluctance machines:
• More analysis and research has to be conducted to find an empirical
or a mathematical relation between the switching frequency of the
switched capacitance circuit and the various parameters of the
resulting current profile, such as the rise and fall times, the peak
value, and the average or RMS value.
• Also, It is desired to undertake more research on the reduction of the
current harmonic content associated with the insertion of the switched
129
capacitance circuit within the converter in order to keep the THD as
low as possible.
• More research on the DCDS topology has to be conducted to validate
its performance. One suggestion is to use the DCDS to resonate in
parallel with the phase inductance so that it could be expelled out of
the circuit at any instant when that is desired.
• The switched capacitance circuit can be introduced to other converter
topologies of the SRM drivers such as the resonant converter
topology which has been confirmed to show discouraging results in [1]
in order to enhance its performance.
130
References
[1] R. Krishnan, Switched Reluctance Motor Drives : Modeling, Simulation,
Analysis, Design, and Applications. Boca Raton, FL: CRC Press, 2001, pp.
398.
[2] T. J. E. Miller, Electronic Control of Switched Reluctance Machines.
Oxford: Newnes, 2001, pp. 272.
[3] T. J. E. Miller, "Optimal design of switched reluctance motors," IEEE
Trans. Ind. Electron., vol. 49, pp. 15-27, 02/. 2002.
[4] K. Nakamura, T. Ono, H. Goto, T. Watanabe and O. Ichinokura, "A novel
switched reluctance motor with wound-cores put on stator and rotor poles,"
IEEE Trans. Magn., vol.
41, pp. 3919-3921, 2005.
[5] S. Song, W. Liu and Y. Wang, "Modeling, dynamic simulation and control
of a four-phase switched reluctance motor," in 2007 IEEE International
Conference on Control and Automation, ICCA, 2008, pp. 1290-1295.
[6] S. Song, W. Liu and Y. Wang, "Modeling, dynamic simulation and control
of a four-phase switched reluctance motor," in 2007 IEEE International
Conference on Control and Automation, ICCA, 2008, pp. 1290-1295.
[7] J. H. Leong, "Design consideration and implementation of switched
reluctance motor drive," in 2002 Student Conference on Research and
Development. SCOReD2002. Proceedings. Global Research and
Development in Electrical and Electronics Engineering, 16-17 July 2002,
2002, pp. 196-9.
[8] G. Rim and R. Krishnan, "Variable speed constant frequency power
conversion with a switched reluctance machine," in Proceedings of 1994
IEEE Applied Power Electronics Conference and Exposition - ASPEC'94, 13-
17 Feb. 1994, 1994, pp. 63-71.
[9] J. Choi, T. H. Kim, Y. Kim, S. Lim, S. Lee, Y. Kim and J. Lee, "The finite
element analysis of switched reluctance motor considering asymmetric
bridge converter and DC link voltage ripple," IEEE Trans. Magn., vol. 41, pp.
1640-1643, 2005.
131
[10] S. Paramasivam, R. Arumugan, B. Umamaheswari, S. Vijayan, S.
Balamurugan and G. Venkatesan, "Accurate rotor position estimation for
switched reluctance motor using ANFIS," in IEEE TENCON 2003.
Conference on Convergent Technologies for the Asia-Pacific Region, 15-17
Oct. 2003, 2003, pp. 1493-7.
[11] N. H. Fuengwarodsakul, R. W. De Doncker and R. B. Inderka,
"Simulation model of a switched reluctance drive in 42 V application," in
IECON'03. 29th Annual Conference of the IEEE Industrial Electronics
Society, 2-6 Nov. 2003, 2003, pp. 2871-6.
[12] D. N. Essah and S. D. Sudhoff, "An improved analytical model for the
switched reluctance motor," IEEE Trans. Energy Convers., vol. 18, pp. 349-
356, 2003.
[13] B. P. Loop and S. D. Sudhoff, "Switched reluctance machine model
using inverse inductance characterization," in Proceedings of 2002 IEEE
Industry Applications Society Annual Meeting, 13-18 Oct. 2002, 2002, pp.
13-21.
[14] J. H. Chen, K. T. Chau, Q. Jiang, C. C. Chan and S. Z. Jiang, "Modeling
and analysis of chaotic behavior in switched reluctance motor drives," in
Proceedings of the 2000 Power Electronics Specialist Conference, 18-23
June 2000, 2000, pp. 1551-6.
[15] K. T. Chau and J. H. Chen, "Analysis of chaotic behavior in switched
reluctance motors using current hysteresis regulation," Electric Power
Components and Systems, vol. 30, pp. 607-24, 06/. 2002.
[16] H. Le-Huy and G. Sybille, "MATLAB/Simulink and PSpice as modelling
tools for power systems and power electronics," in Proceedings of the 2000
Power Engineering Society Summer Meeting, Jul 16-20 2000, 2000, pp. 766-
767.
[17] H. Chen, J. Jiang, S. Sun and D. Zhang, "Dynamic simulation models of
switched reluctance motor drivers," in Proceedings of the 3th World
Congress on Intelligent Control and Automation, Jun 28-Jul 2 2000, 2000,
pp. 2111-2115.
[18] H. Chen, X. Meng and L. Zheng, "Design and simulation software
package of switched reluctance machine systems for windows 9x in english,"
132
in 2000 IEEE Asia-Pacific Conference on Circuits and Systems: Electronic
Communication Systems, Dec 4-6 2000, 2000, pp. 489-492.
[19] Chong-Chul Kim, Jin Hur and Dong-Seok Hyun, "Simulation of a
switched reluctance motors using Matlab/M-file," in 28th Annual Conference
of the IEEE Industrial Electronics Society, 5-8 Nov. 2002, 2002, pp. 1066-71.
[20] G. Gallegos-Lopez, P. C. Kjaer, T. J. E. Miller and G. W. White,
"Simulation study of resonant DC link inverter for current-controlled switched
reluctance motors," in Proceedings of the 1997 2nd International Conference
on Power Electronics and Drive Systems, PEDS. Part 2 (of 2), may 26-29
1997, 1997, pp. 757-761.
[21] K. N. Srinivas and R. Arumugam, "Circuit simulation of dynamic
performances of the switched reluctance motor," in Proceedings of IEEE
Region 10 International Conference on Electrical and Electronic Technology,
19-22 Aug. 2001, 2001, pp. 592-6.
[22] H. Chen, J. Jiang, D. Zhang and S. Sun, "Models and simulation of the
switched reluctance motor drive with the microcomputer control based on
MATLAB software package," in IEEE APCCAS 2000. 2000 IEEE Asia-Pacific
Conference on Circuits and Systems. Electronic Communication Systems.
Proceedings, 4-6 Dec. 2000, 2000, pp. 493-6.
[23] F. Soares and P. J. Costa Branco, "Simulation of a 6/4 switched
reluctance motor based on Matlab/Simulink environment," IEEE Trans.
Aerospace Electron. Syst., vol. 37, pp. 989-1009, 07/. 2001.
[24] O. Ichinokura, S. Suyama, T. Watanabe and H. J. Guo, "A new
calculation model of switched reluctance motor for use on spice," in 8th Joint
Magnetism and Magnetic Materials -International Magnetic Conference-
(MMM-Intermag), Jan 7-11 2001, 2001, pp. 2834-2836.
[25] J. Faiz, J. Raddadi and J. W. Finch, "Spice-based dynamic analysis of a
switched reluctance motor with multiple teeth per stator pole," IEEE Trans.
Magn., vol. 38, pp. 1780-8, 07/. 2002.
[26] J. Mahdavi, G. Suresh, B. Fahimi and M. Ehsani, "Dynamic modeling of
non-linear SRM drive with pspice," in Proceedings of the 1997 IEEE Industry
Applications Conference 32nd IAS Annual Meeting. Part 1 (of 3), Oct 5-9
1997, 1997, pp. 661-667.
133
[27] G. Franceschini, S. Pirani, M. Rinaldi and C. Tassoni, "SPICE-assisted
simulation of controlled electric drives: an application to switched reluctance
drives," IEEE Trans. Ind. Appl., vol. 27, pp. 1103-10, 11/. 1991.
[28] M. G. Giesselmann, "Dynamic modeling of switched reluctance
machines with PSpice for windows," in Proceedings of the 1996 31st
Intersociety Energy Conversion Engineering Conference. Part 1 (of 4), Aug
11-16 1996, 1996, pp. 298-303.
[29] Chang-Chou Hwang and J. N. Lou, "Transient analysis of capacitance
switching for industrial power system by PSpice," Electr. Power Syst. Res.,
vol. 45, pp. 29-38, 04/. 1998.
[30] I. Moson, K. Iwan and J. Nieznanski, "Circuit-oriented model of the
switched reluctance motor for drive systems simulation," in 28th Annual
Conference of the IEEE Industrial Electronics Society, 5-8 Nov. 2002, 2002,
pp. 497-501.
[31] C. Pollock and A. Michaelides, "Switched reluctance drives: A
comparative evaluation," Power Eng J, vol. 9, pp. 257-266, 1995.
[32] J. Oyama, T. Higuchi, T. Abe, K. Haraguchi, E. Yamada and F. Profumo,
"Hybrid type novel switched reluctance motor," in Proceedings of Power
Electronics Specialist Conference - PESC'98, 17-22 may 1998, 1998, pp.
857-63.
[33] P. Rafajdus, V. Hrabovcova, M. Liptak and I. Zrak, "New design of
switched reluctance motor for improving its efficiency," in 16th International
Conference on Electrical Machines. Conference Proceedings, 5-8 Sept.
2004, 2004, pp. 6.
[34] M. A. El-Khazendar and J. M. Stephenson, "Analysis and optimisation of
the 2-phase self-starting switched reluctance motor," in ICEM '86 Munchen.
International Conference on Electrical Machines, 8-10 Sept. 1986, 1986, pp.
1031-4.
[35] Mi-Ching Tsai, Chien-Chin Huang and Zheng-Yi Huang, "A new two-
phase homopolar switched reluctance motor for electric vehicle applications,"
J Magn Magn Mater, vol. 267, pp. 173-81, 12/. 2003.
134
[36] Shang-Hsun Mao and Mi-Ching Tsai, "A novel switched reluctance
motor with C-core stators," IEEE Trans. Magn., vol. 41, pp. 4413-20, 12/.
2005.
[37] Y. Ozoglu, M. Garip and E. Mese, "New pole tip shapes mitigating
torque ripple in short pitched and fully pitched switched reluctance motors,"
Electr. Power Syst. Res., vol. 74, pp. 95-103, 2005.
[38] Jin Woo Lee, Hong Seok Kim, Byung Il Kwon and Byung Taek Kim,
"New rotor shape design for minimum torque ripple of SRM using FEM,"
IEEE Trans. Magn., vol. 40, pp. 754-7, 03/. 2004.
[39] Geun-Hie Rim, Won-Ho Kim, Eun-Soo Kim and Ki-Chul Lee, "A
choppingless converter for switched reluctance motor with unity power factor
and sinusoidal input current," in Proceedings of 1994 Power Electronics
Specialist Conference - PESC'94, 20-25 June 1994, 1994, pp. 500-7.
[40] R. C. Becerra, M. Ehsani and T. J. E. Miller, "Commutation of SR
motors," IEEE Transactions on Power Electronics, vol. 8, pp. 257-263, 1993.
[41] R. Krishnan and P. N. Materu, "Analysis and design of a low-cost
converter for switched reluctance motor drives," IEEE Trans. Ind. Appl., vol.
29, pp. 320-7, 03/. 1993.
[42] A. Hava, V. Blasko and T. A. Lipo, "A modified C-dump converter for
variable reluctance machines," in Conference Record of the 1991 IEEE
Industry Applications Society Annual Meeting (Cat. no.91CH3077-5), 28
Sept.-4 Oct. 1991, 1991, pp. 886-9.
[43] S. Mir, I. Husain and M. E. Elbuluk, "Energy-efficient C-dump converters
for switched reluctance motors," IEEE Transactions on Power Electronics,
vol. 12, pp. 912-921, 1997.
[44] King-Jet Tseng, Shuyu Cao and Jijiu Wang, "A new hybrid C-dump and
buck-fronted converter for switched reluctance motors," IEEE Trans. Ind.
Electron., vol. 47, pp. 1228-36, 12/. 2000.
[45] S. -. Park and T. A. Lipo, "New series resonant converter for variable
reluctance motor drive," in PESC '92 Record. 23rd Annual IEEE Power
Electronics Specialists Conference (Cat. no.92CH3163-3), 29 June-3 July
1992, 1992, pp. 833-8.
135
[46] A. Hava, J. Wacknov and T. A. Lipo, "New ZCS resonant power
converter topologies for variable reluctance machine drives," in Proceedings
of IEEE Power Electronics Specialist Conference - PESC '93, 20-24 June
1993, 1993, pp. 432-9.
[47] Anonymous (1997, Digital signal processing solutions for the switched
reluctance motor. Texas Instruments Europe, Available:
http://focus.ti.com/lit/an/bpra058/bpra058.pdf
[48] Mohamed, A. A. Morsy, "Design and Implementation of Fuzzy Sliding
Mode Controller for Switched Reluctance Motor," 2007.
[49] J. M. Stephenson and J. Corda, "Computation of torque and current in
doubly salient reluctance motors from nonlinear magnetisation data,"
Proceedings of the Institution of Electrical Engineers, vol. 126, pp. 393-6, 05.
1979.
[50] D. Panda and V. Ramanarayanan, "A composite control strategy for
sensorless and low-noise operation of switched reluctance motor drive," in
Proceedings of World Congress on Industrial Applications of Electrical
Energy and 35th IEEE-IAS Annual Meeting, 8-12 Oct. 2000, 2000, pp. 1751-
8.
[51] B. Fahimi, A. Emadi and R. B. Sepe Jr., "Four-quadrant position
sensorless control in SRM drives over the entire speed range," IEEE
Transactions on Power Electronics, vol. 20, pp. 154-163, 2005.
[52] G. Gallegos-Lopez, P. C. Kjaer and T. J. E. Miller, "High-grade position
estimation for SRM drives using flux linkage/current correction model," in
Conference Record of 1998 IEEE Industry Applications Conference. Thirty-
Third IAS Annual Meeting, 12-15 Oct. 1998, 1998, pp. 731-8.
[53] P. C. Kjaer, J. J. Gribble and T. J. E. Miller, "High-grade control of
switched reluctance machines," IEEE Trans. Ind. Appl., vol. 33, pp. 1585-93,
11/. 1997.
[54] A. D. Cheok and N. Ertugrul, "Computer-based automated test
measurement system for determining magnetization characteristics of
switched reluctance motors," IEEE Transactions on Instrumentation and
Measurement, vol. 50, pp. 690-6, 06/. 2001.
136
[55] I. H. Al-Bahadly, "Analysis of position estimation method for switched
reluctance drives," in Proceedings First IEEE International Workshop on
Electronic Design, Test and Applications '2002, 29-31 Jan. 2002, 2002, pp.
262-6.
[56] H. Vasquez, J. Parker and T. Haskew, "Control of a 6/4 switched
reluctance motor in a variable speed pumping application," Mechatronics,
vol. 15, pp. 1061-1071, 2005.
[57] M. T. DiRenzo, M. K. Masten and C. P. Cole, "Switched reluctance
motor control techniques," in Proceedings of the 1997 American Control
Conference. Part 1 (of 6), Jun 4-6 1997, 1997, pp. 272-277.
[58] P. Vas, Artificial-Intelligence-Based Electrical Machines and Drives :
Application of Fuzzy, Neural, Fuzzy-Neural and Genetic-Algorithm-Based
Techniques. UK: Oxford University Press, 1999, pp. 625.
[59] W. Ding and D. Liang, "Modeling of a 6/4 switched reluctance motor
using adaptive neural fuzzy inference system," IEEE Trans. Magn., vol. 44,
pp. 1796-1804, 2008.
[60] B. Mirzaeian, M. Moallem, V. Tahani and C. Lucas, "Multiobjective
optimization method based on a genetic algorithm for switched reluctance
motor design," IEEE Trans. Magn., vol. 38, pp. 1524-1527, 2002.
[61] W. H. Koch, "A Novel Excitation for the Control of Stand Alone Induction
Generators," 1984.
[62] Muhammad Riaz, "Novel Switched Reluctance Motor Drive Systems,"
Thesis, 1985.
[63] BELA, E. F. KARSA, "Analysis of A.C. waveforms harmonic analysis," in
Electrical Measuring Instruments and Measurements Anonymous
BUDAPEST: AKADEMIAI KIADO, 1967, pp. 783-789.
[64] Muhammad H. Rashid, Power Electronics Circuits, Devices, and
Applications. ,3rd ed.USA: Pearson Education International, Printice Hall,
2004, pp. 880.
[65] N. Mohan, Power Electronics Converters, Applications and Design. ,2nd
ed.New York: John Wiley & Sons, Inc., 1995, pp. 802.
137
[66] B. W. Williams, Power Electronics Devices, Drives, Applications and
Passive Components. ,2nd ed.Hong Kong: Macmillan, 1992, pp. 542.
138
Appendix
MATLAB model of SRM
Figure A.1 SRM model.
load
tdev
speed
theta
mechanical
t ime ref sp
loadspeed
Iref
VOLT
Oon
Oof f
THETA
CURNT1
TORQ1
IND1
CURNT2
TORQ2
IND2
CURNT3
TORQ3
IND3
SRM motor
Clock
THETAOFF1
THETAON1
Current
V
139
Figure A.2 Three phases modelling.
9
INDUCTANCE3
8
TORQUE3
7
CURRENT3
6
INDUCTANCE2
5
TORQUE2
4
CURRENT2
3
INDUCTANCE1
2
TORQUE1
1
CURRENT1
In1 Iref
In2 VOLT1
In3 Oon
In4 Oof f
In5 THETA
Out1 CURNT3
Out2 TORQ 3
Out3 IND 3
phase 3
In1 Iref
In2 VOLT1
In3 Oon
In4 Oof f
In5 THETA
Out1 CURNT2
Out2 TORQ 2
Out3 IND 2
phase 2
In1 Ref erence Current
In2 VOLT1
In3 Thetaon
In4 Thetaof f
In5 THETA
Out1 CURRENT1
Out2 TORQUE1
Out3 IND1
phase 1
1
Gain4
THETAIN1
5
THETA
4
Thetaoff
3
Thetaon
2
VOLT
1
Reference
Current
140
Figure A.3 Single-phase model of the SRM.
FLUX
3
Out3
IND1
2
Out2
TORQUE1
1
Out1
CURRENT1
f1
To Workspace12
In1 current1
In2 THETA1
PHASE1 TORQUE
TORQUE 1
In1 Im
In2 Volt
In3 THETA ON
In4 THETA OFF
In5 THETA
In6 i1
f lux
Out1 V
SWITCH 1
MATLAB
Function
MATLAB Fcn
1
s
Integrator
In1 FLUX
In2 THETA1
Out1 CURNT
Out2 IND 1
INDUCTANCE 1
R
Gain
5
In5
THETA
4
In4
Thetaoff
3
In3
Thetaon
2
In2
VOLT1
1
In1
Reference
Current